Abstract
#Underlined indicates abstract presenter.
A multidimensional Phase I trial of an upper limb motor intervention in the acute stroke setting: a novel protocol to investigate dose
1University of Melbourne, Melbourne, Australia. 2Austin Health, Melbourne, Australia. 3Royal Melbourne Hospital, Melbourne, Australia. 4Monash University, Melbourne, Australia. 5Alfred Health, Melbourne, Australia. 6Florey Institute of Neurosciences and Mental Health, Melbourne, Australia
Stroke
Splitting the Difference: Split-Belt Treadmill Training Improves Spatial and Temporal Gait Symmetry in People with Multiple Sclerosis
Colorado State University, Fort Collins, USA
Multiple sclerosis (MS) is a neurodegenerative disease affecting over two million people worldwide.1 This disease is characterized by degradation of the myelin sheath resulting in impaired neural communication throughout the body.2 As a result, most people with MS (PwMS) experience significant mobility impairments. In particular, gait asymmetries between the two legs are prevalent leading to an increased risk of falls, musculoskeletal injury, and decreased quality of life.3 Recent work indicates that split-belt treadmill training, where the speed of each leg is controlled independently, can acutely decrease gait asymmetries for other neurodegenerative impairments.4,5 In this ongoing study, 26 PwMS have undergone a split-belt treadmill training protocol, with the faster paced belt moving under the more affected limb. Gait symmetry is analyzed for each participant at multiple time points during the 15-minute split-belt training protocol using three-dimensional motion capture, force plates, and inertial sensors. Phase coordination index (PCI), which measures temporal coordination to assess stepping accuracy and consistency6, and limb excursion asymmetry (LEA), which measures spatial coordination by quantifying sagittal displacement from toe off to heel strike of the ipsilateral leg7, are the primary outcome measures used to assess gait symmetry in the current study. We hypothesized that participants with a low baseline PCI (i.e., better temporal symmetry) would experience a floor effect, demonstrating less gait adaptation following split-belt treadmill training. Previous work has identified an average PCI for PwMS to be 5.19%, thus we grouped participants based on their baseline PCI value as either 1) above (poor baseline symmetry) or 2) below (good baseline symmetry) the previously recorded MS mean PCI of 5.19%, with the hypothesis that participants with a baseline PCI greater than 5.19% would show a greater response to split-belt treadmill training. Preliminary results show a mean PCI change of -1.31% (SE: 0.65) for the poor baseline symmetry group and a mean PCI change of 0.92% (SE: 0.32) for the good baseline symmetry group with an effect size (d) of -1.15 between groups (p = 0.0062). Among participants with poor baseline symmetry, the mean LEA change is -13.02mm (SE: 5.25) compared to a mean LEA change of 0.83mm (SE: 3.35) for participants with good baseline symmetry. The difference between LEA groups has an effect size (d) of -0.84 and a p-value of 0.037. Our preliminary findings demonstrate an improvement in gait symmetry following spilt-belt treadmill training indicated by a lower PCI and LEA, with the greatest change exhibited from those who have worse baseline measures for each of these outcomes. These findings suggest that PwMS retain the ability for gait adaptation, and provide the template for a novel, targeted mobility intervention.
Multiple Sclerosis (MS)
The Transcallosal Highway: The ipsilateral silent period as a neural biomarker for impaired corpus callosum communication in persons with multiple sclerosis
Colorado State University, Fort Collins, USA
Multiple sclerosis is a neurodegenerative disease that damages the myelin sheath within the central nervous system. This axonal demyelination drastically impacts communication between the brain’s hemispheres in persons with multiple sclerosis (PwMS). Changes in transcallosal communication substantially impairs the coordination of movements that require precise temporal and spatial components such as lower extremity function during gait. These complex bilateral movements require constant communication across the corpus callosum to excite and inhibit specific muscle groups for the desired task. The ipsilateral silent period (iSP) is an indirect marker of the magnitude for transcallosal inhibition in both upper and lower extremities. We hypothesize that the iSP may also serve as a neural biomarker for transcallosal impairments originating from the more affected hemisphere and highlight an underlying mechanism for gait asymmetries in PwMS. Our ongoing study utilizes transcranial magnetic stimulation to assess the inhibitory capacity between the brain’s hemispheres (i.e., iSPs). We analyze muscle activity through electromyography while a focal magnetic pulse is delivered to the first dorsal interosseous. This stimulation results in a suppression of muscle activity reflecting inhibitory transcallosal communication and responses are measured for both hands. There is a lack of research analyzing directionality data between the more and less affected hemisphere in PwMS. Therefore, our results focus on evaluating outcome metrics dependent upon the individual’s more affected hemisphere. Twenty-three PwMS have completed the ongoing protocol and inhibitory metrics such as depth iSP% average, duration, depth iSP% max, transcallosal conduction time, and iSP onset latency were collected. No statistically significant differences have been found between the two hemispheres within PwMS. However, beyond directionality data for transcallosal communication, our study is also investigating gait coordination with overall inhibitory capacity. We will quantify coordination between legs using a metric called the phase coordination index (PCI). A greater PCI value reflects poorer gait coordination. We hypothesize those who demonstrate better gait coordination (measured by lower PCI values), will have a greater iSP. These findings will determine the potential of iSPs as a neural biomarker to address gait asymmetries and stratify participants into the most appropriate mobility rehabilitation protocols.
Multiple Sclerosis (MS)
Potential of High-Definition Transcranial Direct Current Stimulation to Reduce Sensorimotor Impairments Post Hemiparetic Stroke: A Pilot Trial
1University of Oklahoma, Norman, USA. 2University of Oklahoma Health Sciences Center, Oklahoma City, USA. 3UT Health Huston, Huston, USA. 4Northwestern University, Chicago, USA
Stroke is the leading cause of serious long-term disability in the United States. Upper-extremity motor impairments include muscle weakness, abnormal muscle synergies, and spasticity. These impairments occur due to damage to corticofugal tracts and an associated maladaptive upregulation of the cortico-reticulospinal tract. Related to this, the motor and somatosensory cortical regions may reorganize with an increased activity in the contralesional sensorimotor cortices (ipsilateral to the paretic side). Recent studies have demonstrated that transcranial direct current stimulation (tDCS) can be a potentially effective treatment for stroke rehabilitation. However, conventional tDCS is limited by spatial resolution and lack the ability to precisely target a specific brain region. To improve its spatial resolution, this study used targeted high-definition tDCS (HD-tDCS) navigated by paired-pulse transcranial magnetic stimulation. In this double-blind randomized crossover clinical trial, chronic stroke participants (n=5) had three visits 1) anodal HD-tDCS stimulation of the arm region of the primary motor cortex (M1) to improve the function of the corticofugal tracts in the lesioned hemisphere, 2) cathodal stimulation of the arm region of the dorsal premotor (PM) cortex to inhibit maladaptive use of the cortico-reticulospinal tract in the contralesional hemisphere, and 3) sham stimulation. The results demonstrate that anodal (change (mean ± std.): 9.0 ± 7.4) and cathodal (change: 7.2 ± 6.2) stimulation increased Fugl-Meyer Upper Extremity score compared to sham stimulation (change: 1.6 ± 1.5). Additionally, the Erasmus Modification to the Nottingham Sensory Assessment Score improved in one subject after anodal (change: 8.0) and in two subjects after cathodal (change: 3.0 ± 0). However, there is no consistent effect of the HD-tDCS over cortical motor areas on improving sensory functions. These results indicate that both anodal and cathodal HD-tDCS have the potential to decrease paretic upper limb motor impairment following hemiparetic stroke.
Stroke
Effects of priming tDCS expectations on motor learning
Arizona State University, Tempe, USA
Placebo effects are psychophysiological responses to perceived medical treatment. We have recently demonstrated a placebo effect of tDCS on both cognitive and motor training, likely driven by participants’ perceptions of whether they received stimulation and their expectations about the efficacy of tDCS. This study now aims to determine the extent to which reading positive or negative information about the effects of tDCS influences the placebo effect.
Nineteen participants (mean±SD age: 18.6±1.0 years; 7 females) were randomly assigned to read a positive (n=11) or negative (n=8) prompt about the effects of tDCS. We first assessed participants’ de novo expectations about how much tDCS could enhance motor learning. Participants then read positive or negative information about the efficacy of tDCS, followed by another assessment of their expectations of tDCS. Participants then completed 30 trials of a functional upper extremity motor task during which double-blind sham tDCS was applied to the right primary motor cortex (C4). Following stimulation, expectations were assessed again. To ensure blinding, participants were asked at the end of the study whether they believe they received active or sham stimulation. Since participants were reading suggestive information about tDCS, we also administered the Multidimensional Iowa Suggestibility Scale Short Suggestibility Survey to quantify participants’ overall suggestibility. Our primary dependent variable was trial time, a measure of motor task performance where lower values are better and an improvement in trial time indicates learning. We considered the logarithm of trial number (i.e., log trial) in our analysis, given that the rate of task improvement over time was not linear (consistent with motor learning research in general).
As expected, there was a significant effect of log trial on motor performance (p<0.001), indicating performance improved with practice (i.e., learning). There is a trend towards an interaction between log trial and prime on motor performance (p=0.08), although the negative prime group tended to improve more than the positive prime group, contrary to our hypothesis. However, regardless of prime group, higher expectations after priming (p=0.04) and after training (p=0.0005) were associated with more motor learning. This may have been due to participants’ suggestibility, as higher suggestibility scores were associated with more motor learning as well (p<0.001). Lastly, there was a significant effect of perceived stimulation on motor learning (p=0.01), where participants who believed they received active tDCS improved more than those who believed they received sham. This exploratory study provides further evidence of a placebo effect of tDCS on motor training, likely driven by participants’ expectations about the efficacy of tDCS, suggestibility, and belief of whether they received stimulation. Future studies should consider the placebo effects of tDCS and identify their underlying mechanisms to leverage them in clinical care.
Motor Rehabilitation
Is the Reticulospinal Tract a Promising Site for Intervention to Improve Mobility Impairments in People with Multiple Sclerosis?
Colorado State University, Fort Collins, USA
The prevalence of Multiple Sclerosis (MS), a debilitating neurological disease, has increased worldwide since 2013 and currently affects just under 1 million people in the United States. Damage to the central nervous system is a primary corollary of MS and commonly results in mobility and balance impairments, placing people with MS at greater risk of experiencing falls and other adverse events. Identifying functional and structural correlates that can advance our understanding of MS disease mechanisms has the potential to shine light on fruitful rehabilitation efforts. To date, neuromuscular research and rehabilitative interventions aimed at addressing mobility and balance impairments have primarily focused on the corticospinal tract. However, evidence from lesion studies in animals suggests that the reticulospinal tract, a bilateral motor neuron pathway originating in the brainstem, is critical for gross motor movements such as posture and locomotion. More recent literature suggests that resistance training may upregulate reticulospinal tract pathways and present a feasible rehabilitative option. The present study uses transcranial magnetic stimulation as a physiologic measure of neural communication and diffusion tensor imaging as an anatomic measure of neural structure to determine how the reticulospinal tract influences mobility in people with MS and if it presents a promising area for intervention.
Multiple Sclerosis (MS)
Cortical transcranial direct current stimulation influences lower limb cutaneous reflexes in individuals with stroke
University of Illinois Chicago, Chicago, USA
Stroke
Treatment Patterns and Healthcare Costs Among Patients With Stroke and Spasticity
1IQVIA, Falls Church, USA. 2Formerly of IQVIA, Falls Church, USA. 3Ipsen, Cambridge, USA
Peripheral Nerve/Plexus/Neuromuscular Diseases
Switching Adults With Spasticity From OnabotulinumtoxinA to AbobotulinumtoxinA: Real-World Data Across Three US-Based Centers
1Real World Evidence, Cerner Enviza, Malvern, USA. 2Beaumont Health, Royal Oak, USA. 33The Center for Tone Management of the Main Line, Paoli, USA. 4OrthoNeuro, New Albany, USA. 5Ipsen, Cambridge, USA
Peripheral Nerve/Plexus/Neuromuscular Diseases
Neural Mechanisms of Psychomotor Impairment in Adults with Type 1 Diabetes
Colorado State University, Fort Collins, USA
Other
Validating a modified version of the Early Social Communication Scale for assessment of joint attention in infants with visual impairment
1Children’s Hospital Los Angeles, Los Angeles, USA. 2University of Southern California, Los Angeles, USA
Cognitive/Language Rehabilitation
Actual versus predicted values of step length and peak anterior ground reaction force in people post-stroke walking at different gait speeds
1University of Southern California, Los Angeles, USA. 2Johns Hopkins University, Baltimore, USA
Gait dysfunction is common post-stroke. This includes limitations in walking activity and impairments in gait kinetics and kinematics. Clinical practice guidelines recommend gait training at fast speeds to address walking activity limitations post-stroke (1). Importantly, fast walking can also improve kinematic impairments relative to one’s habitual gait pattern post-stroke (2). However, we do not understand how the speed-dependent biomechanical changes achieved by an individual post-stroke compare to those that would be observed if the individual had a neurotypical gait pattern. We recently developed a prediction model (trained on neurotypical gait data) that uses age, leg length, body mass, and gait speed to predict the neurotypical magnitudes of step length and peak anterior ground reaction force (3). Here, we use this model to predict these metrics for people post-stroke walking at different gait speeds. Based on prior work, we hypothesized that the difference between the actual and predicted values of these metrics would increase at faster gait speeds (i.e., post-stroke gait would become increasingly divergent from the predicted neurotypical gait as gait speed increased). We analyzed biomechanical data from the lower extremities of 19 people > 6 months post-stroke (mean age: 63 ± 8 years) who walked on a treadmill at up to three speeds. For each participant, we predicted the values of step length and peak anterior ground reaction force that would be observed in a neurotypical adult of the same age, body mass, leg length, and gait speed. We then calculated the normalized difference between the predicted and the actual values of these gait parameters for the paretic and non-paretic limbs. Linear mixed-effects models were used to evaluate the effects of speed on the difference between the predicted and the actual values of step length and peak anterior ground reaction force on the paretic and non-paretic sides. Random intercepts were included to account for repeated measures. We found no speed-dependent effect on the magnitude of the difference between actual and predicted values of the paretic step length (b= -2.00, P=0.4) and peak anterior ground reaction force (b= -2.08, p=0.6). There was no effect of speed on the difference in peak anterior ground reaction force of the non-paretic limb (b= -0.75, p=0.9), but the percent difference between the actual and predicted non-paretic step length increased significantly with faster walking (b= 5.00, p=0.04). These results suggest that increasing gait speed does not exacerbate impairment in propulsive force generation but may exacerbate impairment in step length. Importantly, this study also demonstrates that we can use a prediction model to estimate the degree of gait impairment without the need for control data.
Stroke
Personalized whole-brain activity patterns predict corticospinal tract activation in real-time
University of Texas at Austin, Austin, USA
The corticospinal tract is the major descending pathway responsible for voluntary human hand movement. Stroke-related lesions that disrupt this tract produce hemiparesis, which interferes with daily life activities. However, many stroke survivors have residual corticospinal projections, the presence of which are strong prognostic markers of poststroke motor recovery. Strengthening these projections could promote poststroke hand motor recovery. Recent work in healthy adults showed that TMS interventions induce long-term potentiation-like plasticity at corticospinal tract synapses and enhance motor learning only when delivered during brain activity patterns reflecting strong corticospinal tract activation (i.e., brain state-dependent TMS). These studies delivered TMS during the same brain activity patterns in all individuals. Because individual stroke survivors have unique patterns of brain damage and motor impairment, poststroke brain state-dependent TMS interventions cannot be delivered during “one-size-fits-all” brain states. Instead, TMS must be delivered during personalized poststroke brain activity patterns reflecting strong residual corticospinal tract activation (i.e., personalized strong corticospinal states). As first steps towards this goal, we developed a novel machine learning-based EEG-TMS system that delivers stimulation during personalized strong corticospinal states in real-time.
Healthy adults completed a single session involving 600 single TMS pulses to the right motor cortex (M1) at 120% of resting motor threshold (RMT) during 64-channel EEG and left first dorsal interosseous EMG recordings. This dataset was used to train personalized machine learning classifiers to identify strong and weak corticospinal states, defined as those EEG activity patterns during which TMS evoked either large or small motor-evoked potentials (MEPs). We spatially filtered whole-brain EEG data (500 ms pre-stimulus segments) and calculated power spectral features (4-35 Hz). Features were ranked according to their predictive ability using the chi-square method. We optimized each participant’s classifier using linear discriminant analysis with 5-fold cross validation, focusing only on high confidence predictions. The best performing classifier during grid search with the fewest features and highest regularization rate was chosen for real-time analysis. We used this classifier to detect strong and weak corticospinal states in real-time and delivered single-pulse TMS (120% RMT) during these states. Preliminary data (N=6) showed that classifier performance during grid search was 74 ± 6%. In each participant, MEPs elicited during strong corticospinal states were larger than those elicited during weak states. MEP amplitudes were 45 ± 20% larger during strong versus weak corticospinal states.
Our results show that personalized power spectral features extracted from whole-brain EEG activity patterns predict corticospinal tract activation in real-time. The magnitude of MEP amplitude modulation observed in this study exceeds that reported using other real-time TMS approaches that deliver TMS during predefined sensorimotor rhythm phases. Overall, our findings represent the first step towards using personalized brain state-dependent TMS interventions to promote poststroke hand motor recovery.
Stroke
A Case Study on the efficacy of beta-blocker eye drops for patients experiencing PCS and TBI symptoms
1Power of Patients, Boston, USA. 2Harbor View Eye Clinic, Portland, USA
More than 80% of patients with post-concussion syndrome (PCS) and traumatic brain injury (TBI) suffer chronic conditions, often with unmet therapies to alleviate their condition impacting the autonomic nervous system (ANS). This dysfunction may be a contributing factor to chronic, persistent PCS and TBI symptoms.
During stress, the adrenal glands produce epinephrine, which increases heart rate and blood pressure, slows digestion, and causes pupil dilation, among other effects. Current research indicates that during TBI and PCS induces ANS dysfunction, causing an overproduction of epinephrine which locks patients into a state of chronic sympathetic arousal. The excess epinephrine binds to β2 receptors on the ciliary body—the muscle which allows the lens of the eye to change shape when focusing on different distances—and prevents it from contracting.
Hyperactive adrenal glands cause an over-production stress hormones locking patients into a “freeze” mode of the flight or fight response mechanism. Working from this hypothesis we theorized that beta blocker drops could be a viable adrenal-stressor treatment for symptoms of TBI and PCS thereby reducing these symptoms allowing patients to return to work, or school, and start vision therapy sooner.
Beta-blockers are competitive antagonists for epinephrine and are therefore able to counteract the effects of the increased epinephrine production by binding to and blocking epinephrine receptor sites on the ciliary body. When the epinephrine is unable to bind, the ciliary body is now able to contract and focus on near tasks; the “brake” has been removed.
The Edinger Westphal (EW) nucleus, which is responsible for pupil constriction and, via cholinergic stimulation, triggers the contraction of the ciliary body. When this area is damaged, the ciliary body may not contract even once the “brake”, or excess epinephrine, is removed3. An accelerator is needed.
Traumatic Brain Injury (TBI)
Does Stimulus Intensity Affect the Ability to Condition Brain Responses and the Associated Short-term Neural Adaptations in Individuals with Anterior Cruciate Ligament Reconstruction?
1School of Kinesiology, University of Michigan, Ann Arbor, USA. 2Department of Physical Medicine and Rehabilitation, Michigan Medicine, Ann Arbor, USA. 3Biomedical Engineering, University of Michigan, Ann Arbor, USA. 4Michigan Robotics Institute, University of Michigan, Ann Arbor, USA. 5Department of Orthopaedic Surgery, Michigan Medicine, Ann Arbor, USA
Although corticospinal excitability contributes to quadriceps dysfunction after anterior cruciate ligament (ACL) reconstruction, current rehabilitation programs do not directly target the corticospinal pathway. Operant conditioning of motor evoked potentials (MEPs) is a promising approach to improve corticospinal excitability; however, the effect of stimulus intensity on operant conditioning of MEPs and the feasibility of this approach in individuals with ACL reconstruction have not been well studied. Therefore, this study evaluated 1) if corticospinal excitability can be improved via operant conditioning in a single session and whether up-conditioning can result in short-term improvements in neural excitability and 2) does stimulus intensity used during operant conditioning affect these outcomes in individuals with ACL reconstruction. Thirty-six individuals with ACL reconstruction (12.4 ± 7.7 months post-operative) were trained to increase their transcranial magnetic stimulation (TMS)-induced MEPTORQUE of the quadriceps muscles using an operant conditioning paradigm. Participants were randomized into one of three groups based on the participant’s active motor threshold (AMT) (100%, 120%, and 140% AMT). Participants performed operant up-conditioning procedures for a total of 225 training trials (3 blocks of 75 trials) while seated on a dynamometer with their reconstructed leg fixed at 60° of knee flexion. MEPTORQUE responses were elicited using a 110 mm double cone coil attached to a Magstim 2002 stimulator. TMS recruitment curves were collected at several intensities (100%, 110%, 120%, 130% and 140% of AMT]) while the subject maintained a background contraction (10% of maximum). A block of 20 control trials (i.e., no up-conditioning) was also collected immediately before (BASE) and after up-conditioning procedures. The improvements in peak-peak MEPTORQUE between the baseline control and the conditioning (COND) blocks were used to evaluate the feasibility of up-conditioning. The area under the curve (AUC) of the MEPTORQUE obtained during the recruitment curves before and after up-conditioning were used to evaluate the short-term adaptations in neural excitability. Linear mixed models were used to evaluate the changes in MEPTORQUE during (feasibility) and after operant conditioning (neural adaptations) and the effect of TMS intensity on these outcomes. Results indicated that individuals with ACL reconstruction were able to up-condition their MEPTORQUE in a single session (p<0.001; BASE: 17.27 ± 1.28, COND: 21.35 ± 1.28 [mean ± standard error (SE)]), but this ability was not affected by the stimulus intensity used during training (p=0.841). Similarly, a single session of operant conditioning resulted in significant improvements in neural excitability (p=0.047; pre: 687.91 ± 50.15, post: 761.08 ± 50.15 [mean ± SE]), and was not affected by stimulus intensity (p=0.669). These findings indicate that operant conditioning may be a feasible approach to improving corticospinal excitability in individuals with ACL reconstruction and any of the three stimulus intensities studied herein could be effectively used in future operant conditioning paradigms.
Motor Rehabilitation
Combined electrical simulation and treadmill training intervention on gait performance in post-stroke individuals
1Neuroscience Program, Loyola University Chicago, Chicago, USA. 2Department of Physical Therapy and Assistive Technology, National Yang Ming Chiao Tung University, Taipei, Taiwan. 3Department of Engineering, Loyola University Chicago, Chicago, USA
Motor Rehabilitation
Tracking walking recovery in individuals with motor incomplete spinal cord injury with transcranial magnetic stimulation: preliminary findings
1Texas Woman’s University, Dallas, USA. 2Baylor Scott & White Research Institute, Dallas, USA
Spinal Cord Injury (SCI)
Are we doing enough: Neurorehabilitation outcomes pertaining to stroke population in an acute inpatient rehabilitation unit
Ochsner Clinic Foundation, New Orleans, USA
Stroke
The role of proprioception in online movement control: Insights from reaching arm movements in a patient with Large Fiber Sensory Neuropathy
1University of Minnesota, Minneapolis, USA. 2Pennsylvania State University, State College, USA. 3Pennsylvania State University College of Medicine, Hershey, USA. 4Aix Marseille Université, CNRS, ISM, Marseille, France
Proprioception provides crucial information that is necessary for the nervous system to efficiently coordinate the limbs and produce accurate movement. Previous studies with individuals who are deprived of proprioception, due to a massive Large Fiber Sensory Neuropathy, have shown that they are unable to properly update their internal models of movement, resulting in movement deficits (Sainburg et al. 1993, 1995). These patients tend to use vision to partially compensate for the absence of proprioception to make accurate reaching movements (Blouin et al. 1993). However, it remains unclear whether online responses to visual perturbations rely on proprioceptive feedback due to mixed findings in the literature. For example, Bard et al. (1999) reported that corrective movements in response to a small target displacement were unimpaired in a deafferented patient (GL), while Sarlegna et al. (2006) observed abnormal corrections in response to large target displacements in the same patient. With respect to a perturbation of the cursor (i.e., the visual representation of hand position), Bernier et al. (2006) showed that GL could adapt as well as controls to a systematic perturbation. Our primary interests in the current study are with respect to online response, rather than adaptation, to the perturbations. We predicted that in the absence of proprioception, target displacements should have less impact on performance than cursor displacements because such perturbations do not imply a change in limb position. The latter should affect estimates of limb position, while the former should not. We recruited patient GL (right-handed; 70 years old; 40 years after the onset of peripheral deafferentation) and 15 right-handed age-matched female control participants. The experiments consisted of a target displacement task and a cursor displacement task designed on the Kinereach virtual reality motion tracking system. Each task consisted of 150 trials completed with each hand separately. There were 6 possible lateral displacements of the cursor position or target position, and would occur pseudorandomly at movement onset. Preliminary data analyses show that both types of displacements produced significantly larger performance errors at the end of trial in GL compared to controls, highlighting the role of proprioception in the online control of visually-guided movements.
Peripheral Nerve/Plexus/Neuromuscular Diseases
A Review of Disparities in Racial and Ethnic Inclusion in Stroke Rehabilitation Clinical Trials
1Arizona State University, Tempe, AZ, USA. 2Washington University School of Medicine, St. Louis, MO, USA
Black Americans have a significantly higher risk of stroke when compared to White Americans. However, race is not only a factor in stroke incidence, but it also presents a considerable impact in the physical impairment of Black Americans following stroke occurrence. While it is still unclear as to whether race and ethnicity play a significant role in stroke rehabilitation outcomes, it is imperative to look further into racial and ethnic representations of participants in stroke rehabilitation clinical trials. Thus, the purpose of this review was to quantify racial and ethnic inclusion and reporting trends in stroke rehabilitation clinical trials. Here, race has been defined according to its sociological definition. The Centralized Open-Access Rehabilitation Database for Stroke (SCOAR) database was used to select the trials reviewed in this study. The SCOAR database consists of 215 independent clinical trials around the world from March 1977 to April 2016. Only studies conducted in the U.S. are being considered, primarily because of the National Institutes of Health’s mandate to collect racial and ethnic data in human trials (which was implemented in 1997). At this point, data extraction is still ongoing, but in a preliminary analysis of 36 U.S. trials, 80% did not report any participant racial or ethnic demographics, and 77.3% of trials that did not report racial or ethnic data were funded by the National Institutes of Health. Out of the 7 trials that did provide data on participant race or ethnicity, only 5 reported percentages of different racial or ethnic categories (i.e. Black, Hispanic, White, Asian, etc.) compared to the remainder that simply categorized participants’ race as “white” or “other.” These preliminary results demonstrate a clear need to report racial and ethnic data in stroke rehabilitation trials. Once more data has been extracted, we will test whether racial composition (measured as % White) and ethnic composition (measured as % non-Hispanic) is associated with effect size (measured as Hedges’ g). We acknowledge the real challenges in recruiting a diverse sample within a given trial, but publishing demographic information could allow meta-analytic approaches to collectively consider the effects of race and ethnicity on stroke rehabilitation.
Stroke
Speed-based high intensity interval treadmill training as a measure of intensity post stroke
University of Illinois at Chicago, Chicago, USA
Stroke
Tele-tDCS for ALS: A case series examining safety, feasibility and preliminary effectiveness
1University of Illinois at Chicago, Chicago, USA. 2
Peripheral Nerve/Plexus/Neuromuscular Diseases
Multi-site generalization of clusters of walking impairment in individuals with chronic stroke
1Chapman University, Irvine, USA. 2University of Southern California, Los Angeles, USA. 3Kennedy Krieger Institute and Johns Hopkins, Baltimore, USA. 4Emory University, Atlanta, USA. 5University of Pittsburgh, Pittsburgh, USA
Stroke
A Cross-Device Investigation of the Strength of Placebo Effects of Transcranial Direct Current Stimulation (tDCS) on Motor Training: Comparing HD and Traditional tDCS
Arizona State University, Tempe, USA
Transcranial direct current stimulation (tDCS) is emerging as a potential adjuvant to traditional therapy in motor rehabilitation. However, its ability to enhance motor learning remains controversial. We have previously shown that the amount of improvement in motor learning during online stimulation is associated with how much people expect tDCS to work, indicating a significant placebo effect of tDCS. We have also replicated this finding when tDCS is applied during cognitive training. Regarding tDCS, there are generally two types of devices used experimentally: novel High-definition (HD) tDCS and traditional 1x1 tDCS. The purpose of this study was to compare the magnitudes of placebo effects between these two devices. Because HD-tDCS systems appear more complex and technologically advanced relative to the 1x1 system, we hypothesized that participants’ expectations of the HD-tDCS system would be higher than those of the 1x1 system, resulting in a larger placebo effect on motor learning. Fifty participants (mean±SD age: 21.24±3.47 yrs; 20 females) were randomly assigned in a single-blind fashion to one of four groups: sham 1x1 tDCS (n = 20), control 1x1 tDCS (n = 19), sham HD-tDCS (n = 6), and control HD-tDCS (n = 6). (Note: HD-tDCS data collections are still ongoing). Sham groups were subjected to stimulation in the right primary motor cortex (C4) that ramped up to 2 mA for 30 seconds then back down to 0 mA for 30 seconds during the first and last minute of a 20-minute stimulation session during motor training. Control groups were open-label, as they were shown that the tDCS device was not plugged in but was still placed appropriately on the head to keep any somatosensory effects constant. For motor training, all participants completed 30 trials of a functional upper extremity task. The primary measure of task performance was trial time (i.e., the time to complete each trial, where lower scores indicate better performance). Linear mixed-effects models tested for placebo effects of tDCS on motor performance, dependent on group assignment. As expected, task performance improved with practice, whereby trial time was significantly affected by trial number (<0.0001), but there was no significant difference in placebo effect between groups, relative to the HD sham group’s performance (p>0.05). There was also no difference in tDCS expectations based on device type (p=0.45). Thus, at this time, there are no significant differences in the magnitude of placebo effect based on device type. This suggests that people’s perceptions of the complexity of neurotechnology do not influence any placebo effect of tDCS.
Motor Rehabilitation
The impact of socioeconomic and environmental factors on motor skill acquisition among a nationwide cohort across the lifespan
1Arizona State University, Tempe, USA. 2Oregon Health and Science University, Portland, USA
The ability to acquire motor skill is fundamental to neurorehabilitation. However, there is sparse knowledge about whether environmental and socioeconomic factors (i.e., air quality, lead levels in water, food access, and socioeconomic disadvantage) affect motor skill acquisition. If these factors impact motor skill acquisition, then clinicians and researchers may need to alter interventions to account for these factors. The purpose of this study was to examine if these environmental and socioeconomic variables modified individual motor skill performance while controlling for common demographic variables such as age, sex, race, and education. In the current, online motor skill study conducted through Amazon Mechanical Turk, eligible participants completed a brief demographic and lifestyle survey prior to completing 75 trials of a complex motor skill. Participants also provided their residential zip code, which was used to derive individual environmental and socioeconomic factors from different open-access databases, including the Individual Air Quality index (AQI) from the Environmental Protection Agency, the Area Deprivation Index (ADI) from the Health Resources and Services Administration, lead levels from monitored water sources from the United States Geological Society, and food access from the United State Department of Agriculture. Overall, 878 participants (mean age = 39.52, sd = 11.32, Female = 454) who completed the motor task on their personal internet connected device (e.g., tablet, smartphone, desktop computer) were included in the final analysis. Results demonstrated that participants who reported living in a zip code with more air pollution and greater more socioeconomic disadvantage performed worse on the motor task compared to those with less air pollution and socioeconomic disadvantage. Motor task performance was not affected by lead levels or food access. This study demonstrated how environmental and socioeconomic factors derived from self-reported zip code were associated with motor skill acquisition among a nationwide cohort. Such findings could have implications for clinical care and research in motor interventions. As such, future research should investigate potential causal mechanisms of the observed results, the generalizability of these findings to other types of interventions, and methods for reducing their potential impact in the neurorehabilitation process.
Motor Rehabilitation
Relationship of changes in circulating BDNF and motor impairment following a stroke rehabilitation intervention
Medical University of South Carolina, Charleston, USA
Main Outcome Measure(s): Pre- and post-intervention, participants undertook assessments for; 1) upper extremity impairment, evaluated with the Fugl-Meyer Upper Extremity Assessment [FMA-UE], and 2) blood samples for assessment of serum [S-] and plasma [P-] BDNF. Absolute changes were calculated for all measures and correlational analyses were performed
Stroke
Effect of single session of repetitive transcranial magnetic stimulation applied to different brain regions on balance performance after stroke
Texas Woman’s University, Dallas, USA
Stroke
Effects of anodal tDCS stratified by corticospinal organization on motor excitability in children with hemiparetic cerebral palsy
1Marquette University, Milwaukee, USA. 2Medical University of South Carolina, Charleston, USA. 3University of Wisconsin-Madison, Madison, USA. 4National Institute of Mental Health, Bethesda, USA. 5University of Minnesota, Minneapolis, USA. 6Gillette Children’s, St. Paul, USA
Children with hemiparetic cerebral palsy (HCP) due to prenatal/perinatal stroke experience lifelong impairments in motor function. Transcranial direct current stimulation (tDCS) may be a safe and feasible adjuvant therapy to augment intensive neurorehabilitation, however there is variability in outcomes of published clinical trials incorporating tDCS. Conventional tDCS protocols have focused on either increasing or decreasing the excitability of the ipsilesional or contralesional motor cortex, respectively. Using single-pulse transcranial magnetic stimulation (spTMS), studies demonstrate that the contralesional hemisphere provides inputs to both limbs in about 50% of children with HCP due to neuroplastic changes in corticospinal development following stroke. Thus decreasing the excitability of the contralesional hemisphere may be detrimental for motor system plasticity. Developing tailored stimulation protocols based on individual corticospinal organization patterns may increase the efficacy of tDCS and reduce variability in outcomes. The goal of this study was to examine the safety and immediate effects of a single-session of anodal tDCS on corticospinal excitability wherein stimulation was applied based on individual corticospinal organization.
Fourteen children with HCP (age=13.8 ± 3.63) were stratified into two corticospinal organization subgroups based on spTMS-confirmed motor evoked potentials (MEP) from the abductor pollicis brevis: ipsilesional MEP presence (MEPIL+) or absence (MEPIL−). Stratified subgroups were randomized to receive real or sham anodal tDCS (1.5 mA) for 20 minutes: the anode was applied to the ipsilesional (MEPIL+ group) or to the contralesional (MEPIL− group) motor cortex; the cathode was applied to the contralateral forehead. Participants performed a finger tracking motor task concurrently with tDCS. Safety was assessed with questionnaires and motor function evaluations. Corticospinal excitability was assessed at baseline (pre-tDCS) and every 15 minutes for 1 hour after tDCS (post-tDCS).
No serious adverse events occurred, anticipated minor side effects were reported and self-limited, and no transient declines in hand function were noted. Six of 14 participants had MEPs in the paretic hand during spTMS of the ipsilesional motor cortex (MEPIL+ group). The mean ± SD change in MEP amplitude from the paretic hand immediately following tDCS relative to baseline was +41.7% ± 122.5% (real tDCS) and −14.2% ± 48.2% (sham tDCS). Thirty minutes after tDCS, the change in MEP amplitude was +57.6 % ± 132.3% (real tDCS) and −30.2% ± 30.5% (sham tDCS).
Anodal tDCS of either the contralesional or ipsilesional motor cortex was safe and feasible. While still variable, the effects of anodal tDCS to increase corticospinal excitability were consistent with prior studies in typically-developing participants in both pediatric and adult populations. Altogether, this proof-of-principle study indicates the potential for tailored tDCS protocols for children with HCP. Additional research involving expanded experimental designs is needed to confirm these effects and to determine the feasibility of translating tailored protocols into a clinically-relevant intervention.
Motor Rehabilitation
Mindset, environment, and participation: factors chronic stroke survivors identify as influencing movement behavior and recovery
1University of Southern California, Los Angeles, USA. 2University of Montreal, Montreal, Canada
Stroke
Walking Faster and Carrying More Weight: How Triceps Surae Activity Contributes to Increasing Speed and Bearing Added Weight in Human Locomotion
College of Health Professions, Medical University of South Carolina, Charleston, USA
Motor Rehabilitation
Subthalamic Connectivity in Participants with Parkinson’s Disease and Freezing of Gait
Medical University of South Carolina, Charleston, USA
Other
Individuals with Hemiparetic Stroke Abnormally Perceive their Elbow Torques when Abducting their Paretic Shoulder
1Northwestern University, Chicago, USA. 2Virginia Polytechnic Institute and State University, Blacksburg, USA
Stroke
The tradeoff between kinematic and muscular control of reaching as a potential biomarker of motor performance in stroke
1Georgetown University, Washington, DC, USA. 2The Catholic University of America, Washington, DC, USA
Nearly 3 million Americans live with arm impairment following stroke, and nearly 80% of patients are unable to fully recover. Moreover, individual differences in recovery make one-size-fits-all rehabilitation approaches (i.e., constraint-based therapies) suboptimal. We aimed to create a rehabilitation platform for stroke survivors, utilizing virtual reality and exoskeleton technologies, to examine individual differences in upper extremity neuromuscular control. Patient control characteristics may serve as a treatment biomarker, paving the way for personalized approaches to stroke rehabilitation.
Stroke survivors (N = 11, time since stroke: 3.7 +/- 3.1 years) reached for targets in a virtual reality environment using both hands. They completed 162 reaches divided into 3 blocks. Following baseline, we used our custom exoskeletons to provide 50% arm weight assistance to the impaired limb and 50% arm weight resistance to the non-impaired limb. We removed the exoskeletons during the retention block. We used electromyography to approximate muscle activity in the anterior deltoids. Relative contribution (RC) was calculated as the displacement of the impaired arm divided by the sum of displacements for both arms. Muscle contribution (MC) was calculated as the root mean square of impaired arm muscle activity divided by the sum of activity for both deltoids, normalized to maximum voluntary contraction.
During baseline, RC of the impaired limb was 46%; patients reached significantly less with their impaired arm compared to their non-impaired arm (p = 0.02). MC of the impaired deltoid was 45% and did not differ from 50% (p = 0.57). During loading, RC did not change relative to baseline (p = 0.9), but MC tended to decrease by 9% (p = 0.15). These preliminary results suggest a tradeoff between kinematic and muscular control of reaching. This new finding closely matches our previous work in 12 healthy controls, where we found a 2% increase in RC and a 11% decrease in MC. Importantly, 5/11 patients exhibited an inverse tradeoff (i.e., decrease in RC and/or increase in MC). We will analyze participant neuroimaging data to determine the role lesion size and location play in predicting an individual’s response to gravity compensation.
Our tradeoff analysis serves as a potential biomarker of stroke recovery that integrates both behavioral and neuromuscular control characteristics. Individual control characteristics may form the basis for personalized technologies for stroke rehabilitation. Our work paves the way for graded-constraint therapy, allowing clinicians and researchers to fine-tune constraint levels based on the individual needs of the patient.
Stroke
Movement-related cortical stimulation for enhancing corticospinal excitability below the level of incomplete spinal cord injury: A proof-of-concept case study
1Case Western Reserve University, Cleveland, USA. 2MetroHealth Center for Rehabilitation Research, Cleveland, USA
After sustaining an incomplete spinal cord injury (iSCI), not only is the transmission of signals at the site of the injury affected, but there is evidence that higher brain structures are impacted as well. Studies have shown that the excitability of intracortical circuits within the primary motor cortex is decreased and delayed, leading to decreased ability for initiating movement and recruiting residual spinal motor neurons. Non-invasive brain stimulation, such as transcranial magnetic stimulation (TMS), has been studied as a modality for enhancing corticospinal excitability and facilitating muscle activation below the level of injury in individuals with iSCI. One approach to target intracortical circuits is to pair TMS with motor intention; a technique known as movement-related cortical stimulation. In able-bodied participants, movement-related cortical stimulation has been shown to modulate corticospinal excitability in a spike-time dependent plasticity manner. Here, we hypothesize that delivering TMS during motor intention (i.e. while intracortical networks are facilitated) will enhance corticospinal excitability and thus improve activation of muscles below the level of injury. In a proof-of-concept case study, one person with chronic incomplete tetraplegia (AIS C, sensory C8, motor T1) participated first in a crossover study (Experiment 1), where we investigated the impact of TMS timing on corticospinal excitability when delivered 50 ms prior to or after movement (1-week washout), based on muscle activity of the abductor halluces. In a follow-up experiment (Experiment 2), the participant received 5-consecutive treatment days of movement-related cortical stimulation where active TMS was delivered 50 ms prior to movement onset for 15-20 minutes (120 total stimuli per session). Experiment 1: Corticospinal excitability was assessed before and immediately following each Intervention: (a) Sham TMS delivered 50 ms prior to movement onset; b) Active TMS delivered 50 ms prior to movement onset; c) Active TMS delivered 50 ms after movement onset. We found an increase in corticospinal excitability when TMS was delivered prior to movement initiation (i.e. during motor intention), but not following sham stimulation and when active TMS was delivered after movement onset. Experiment 2: We assessed corticospinal excitability and volitionally controlled motor unit recruitment at baseline, the beginning of each treatment session, post 3-days treatment, and post 7-days treatment. Five consecutive treatment sessions of movement-related cortical stimulation resulted in increased corticospinal excitability as well as improved volitionally controlled motor unit recruitment up to 3-days post treatment, where outcome measures returned to baseline by 7-days post treatment. Future studies will extend these findings to a larger sample population and investigate whether movement-related cortical stimulation can be used to prime the motor circuitry prior to receiving physical and/or occupationial therapy in order to impact motor recovery in the iSCI population.
Spinal Cord Injury (SCI)
Concurrent anodal HD-tdcs to the left, but not the right, posterior-parietal cortex enhances learning and interlimb transfer of a skill task
1Penn State University, University Park, USA. 2Penn State Milton S. College of Medicine, Hershey, USA
Previous evidence indicates that direction learning in a motor task is specialized to the dominant hemisphere/arm. In addition, following visuomotor rotation adaptation, learning transfers asymmetrically between the arms. Emerging evidence from lesion studies and stimulation studies has indicated that the left posterior parietal cortex plays a role in learning such sensorimotor mapping and how such learning transfers between the arms. However, these findings have been limited to adaptation paradigms, in which the task requires learning to move in the presence of a perturbation. We now examine whether the left posterior parietal cortex (PPC) also mediates motor direction learning in a non-perturbation task, and whether this learning is lateralized. We present a virtual air hockey task in which the player controls a paddle to strike a puck toward a target. Participants were instructed to initiate the movement from a starting position located at 45° and 135°relative to and 10 cm distance from the puck. The puck direction was determined as the direction that passes the center of the puck from the location of impact at the periphery of the puck, and the puck distance was determined by the amplitude of the vector component of cursor velocity toward that direction with a virtual friction coefficient. The target was elongated along the goal direction, so participants were required to control puck direction while control of puck distance was minimally constrained. Participants were randomly assigned to three groups: Left PPC stim (LPPCS), Right PPC stim (RPPCS), and sham. They practiced the task initially using the dominant arm concurrent with 20 minutes of 2 milliamp HD-tDCS. Following this, participants used the non-dominant arm to assess inter-manual transfer of learning. Based on our hypothesis that the left-PPC is specialized for representation and mapping of movement direction, we predicted better task learning and inter-manual transfer in the LPPCS group compared to other groups. Our data support this prediction for the early stage of learning. In contrast, the RPPCS group showed no differences from the sham group in learning or transfer. We conclude that motor task direction mapping appears to be lateralized to the left posterior parietal cortex. This finding supports previous evidence from visuomotor rotation studies and expands those results to direction learning in skill-tasks, such as our air hockey game. These findings may eventually be applicable to facilitating responses to rehabilitation in patient populations with impaired motor control.
Motor Rehabilitation
Investigating the Relationship Between Altered Functional Connectivity and Sensorimotor Control in Chronic Stroke
1Medical University of South Carolina, Charleston, USA. 2Ralph H. Johnson VA Health Care System, Charleston, USA
Stroke
Sensory circuits for hand function in pediatric hemiplegia: a bedside to bench study
Columbia University, New York, USA
Sensory Rehabilitation
Protocol of a pilot clinical study evaluating a novel brain stimulation approach to promote bimanual motor function and control in chronic stroke
1Cleveland Clinic Lerner Research Institute, Cleveland, USA. 2MetroHealth Center for Rehabilitation Research, Cleveland, USA. 3Case Western Reserve University, Cleveland, USA. 4Cleveland Clinic Imaging Institute, Cleveland, USA. 5Cleveland Clinic Neurological Institute, Cleveland, USA
The majority of stroke survivors have chronic upper limb paresis that limits their ability to perform daily tasks.1 Most daily tasks require bimanual arm and hand use.2 Superior bimanual function is associated with better functional independence after stroke.3 Yet, bimanual motor function is underemphasized in stroke rehabilitation research and interventions generally focus on promoting unimanual gains. Unfortunately, unimanual gains do not fully transfer to improved bimanual abilities.4 Here, we seek to test the hypothesis that non-invasive brain stimulation delivered to contralesional higher motor cortices (cHMC) using repetitive transcranial magnetic stimulation (rTMS) when combined with bimanual motor training will lead to favorable effects on bimanual motor function and control in chronic stroke survivors. CHMC are anatomically specialized to control bimanual movement given their dense interhemispheric, crossed and uncrossed connections.5–7 A single session of cHMC rTMS led to greater improvements on bimanual elbow flexion coordination than a single session of rTMS given to ipsilesional primary motor cortex.8 Our ongoing triple-blinded pilot randomized controlled study will evaluate feasibility and estimate the effects of delivering multiple sessions of cHMC rTMS in combination with bimanual rehabilitation in persons with chronic stroke. Participants 18-90 years of age who are >6 months post unilateral ischemic or hemorrhagic stroke are included. Individuals are required to retain ≥10º wrist/finger/thumb extension but score ≤11/14 points on the hand section of upper extremity Fugl-Meyer (UEFM). Participants are randomized to receive cHMC rTMS or sham rTMS immediately preceding rehabilitation 2x/week for 6 weeks. Rehabilitation consists of 100% bimanual functional task practice. Assessments of bimanual motor function and control, neurophysiology and resting-state functional connectivity are made twice at baseline and repeated at end-of-treatment. Motor function and control tests are also repeated at 1-month follow-up. Bimanual motor function is assessed using the Bimanual Assessment Measure (BAM, primary outcome), a valid and reliable test for bimanual function in stroke.9 Bimanual motor control is assessed as ability to control and modulate bilateral grip forces to reach a common goal and independent goals (symmetrical and asymmetrical). Primary and secondary outcome measures will be analyzed using 2-way repeated measures ANOVA (GROUPxTIME). Association between motor gains, neurophysiologic and imaging-based changes will be studied. To the best of our knowledge, this is the first study to date to evaluate effects of neural priming given using brain stimulation to promote effects of bimanual motor training. This study started in April 2022 and is expected to complete enrollment by December 2023. Baseline data from 4 participants will be presented to demonstrate methodology and approach. Repeat baseline design will allow evaluation of test-retest reliability and characterization of minimal detectable change (MDC). Effect size estimates will be generated to determine sample sizes for future investigation of bimanual function in stroke.
Motor Rehabilitation
Ischemic conditioning to improve motor and neurophysiological outcomes post-stroke: a scoping review
University of Illinois Chicago, Chicago, USA
Current stroke rehabilitation techniques to improve motor function are investigating adjunct therapeutic strategies, substantiated in neurophysiological mechanisms, to augment standard training methods. Despite promising outcomes, no modality has emerged as the most effective to stimulate the desired changes in the central nervous system for improvements during rehabilitation. Hence, there is a critical need to investigate alternative adjunct therapies to improve physical function post-stroke, especially for those who are unaffected by current supplemental practices. Ischemic conditioning (IC), a procedure where the limb is exposed to brief, repeated bouts of blood flow occlusion immediately followed by reperfusion, has high clinical relevance for stroke rehabilitation but is in its relative infancy. IC is based on another procedure known as blood flow restriction training which combines blood flow restriction with low-intensity exercise to echo the effects of high-intensity exercise. This technique has been used extensively in healthy and clinical populations for individuals who are typically unable to execute traditional exercise protocols. Despite the potential benefits of IC, our knowledge on the application and efficacy of IC in stroke is limited. We performed a scoping review to synthesize the current evidence regarding neurophysiological and motor effects of IC, with or without exercise, in stroke. Moderate to strong evidence from six studies were included in this review. Studies by Durand et al. (2019), Feng et al. (2019), and Hyngstrom et al. (2020) investigated the chronic effects of IC in stroke by administering stationary intermittent IC anywhere from 2 weeks to 6 months. These studies demonstrated improvements in walking speed, fatiguability, cognition, and endothelial function. Hyngstrom et al. (2018) explored the acute effects of intermittent IC and found improvements in paretic leg strength and motor unit behavior. One study by Kjeldson et al. (2022) investigated the neurophysiological effects of continuous IC during lower limb resistance exercise, but results were inconclusive as researchers were unable to obtain motor responses from the target limbs for a significant number of participants. An additional study by Du et al. (2021) examined the effects of continuous IC during whole body resistance exercise and found significant increases in BDNF, VEGF, and blood lactate concentration that were comparable to high-intensity exercise. Our scoping review suggests that IC is a promising clinical approach in stroke, however as demonstrated by the included studies, there are protocol inconsistencies that need to be addressed before further implementation into stroke rehabilitation. Parameters to be addressed include the number of sessions, time length of IC application, continuous versus intermittent IC cycles, cuff pressure, and remote versus local cuff placement. High quality studies with more standardized procedures focusing on neurophysiological mechanisms are required to establish the efficacy and underlying mechanisms of IC in stroke.
Stroke
Non-Primary Motor Area Involvement in Reaching Behavior After Stroke
1University of Pittsburgh, Pittsburgh, USA. 2VA Pittsburgh Healthcare System, Pittsburgh, USA
Stroke is a major cause of disability, motivating the need for better therapies. Motor impairments following a stroke can be a source of decreased independence and quality of life for patients. Our overall goal is to improve motor function after stroke using a combination of neuromodulation and robot-guided practice. To do so, we must first understand the neural mechanisms underlying essential daily movements, particularly reaching. Previous literature shows that non-primary motor areas are involved in different aspects of reaching, but how their connections to M1 change functionally after stroke remains unclear. In this study, we investigate premotor and parietal connectivity during reaching by disrupting these areas with a repetitive transcranial magnetic stimulation (rTMS) protocol in stroke and healthy participants. The results could provide the foundations for neuromodulatory strategies to maximize therapeutic outcomes.
We recruited a preliminary cohort of 2 subcortical stroke patients and 9 healthy controls. Participants performed a simple planar reaching task with their right or impaired hand while seated in a robotic exoskeleton. They must make a forward or backward reach to a target within 1.5s of an audio go cue, after which visual feedback of the target and fingertip cursor disappear. We delivered triple-pulse rTMS 100ms before the reaction time over 6 locations: left and right dorsal premotor cortex (LPMd and RPMd), left and right ventral premotor cortex (LPMv and RPMd), left and right posterior parietal cortex (LPPC and RPPC), and the post-central sulcus (PCS) (control region). Pulses were administered at three intensities: 80% resting motor threshold (RMT), 120% RMT, and no stimulation (control). 36 reaches were completed at each stimulation location with the three intensities randomly interspersed. Go cue timings and reach directions were also randomized.
Kruskal-Wallis tests were performed with Dunn’s post-hoc tests and Bonferroni correction to evaluate the effects of stimulation location and intensity on kinematic metrics. In some healthy controls, LPMd stimulation at 120% RMT increased the total path length ratio and distance from straight line at maximum velocity and decreases the path length ratio at maximum velocity. This suggests longer, less efficient reaches with velocities that peak earlier in the reach. LPMv stimulation at 80% also increased endpoint error. In summary, disrupting premotor regions can hinder reaching in healthy controls. In one stroke patient, stimulation resulted in no effects. In the other, LPMv stimulation at 120% RMT appeared to decrease endpoint error while RPMv and RPPC stimulation at 80% increased the path length ratio at maximum velocity, indicating slower but more accurate reaches. This improvement in reach accuracy shows promise in reinforcing the connectivity of non-primary motor areas. Continued work will move towards developing a treatment that synchronizes stimulation with practice.
Stroke
Operant Conditioning of the Soleus Cutaneous Reflex in a Person with Chronic Incomplete Spinal Cord Injury: Implications on Pain Perception
Medical University of South Carolina, Charleston, USA
Spinal Cord Injury (SCI)
Proprioceptive Thresholds as a Potential Predictor of Sensorimotor Function After Stroke
1University of Delaware, Newark, DE, USA. 2Queen’s University, Kingston, Ontario, Canada. 3University of Calgary, Calgary, Alberta, Canada
Sensorimotor impairments are common after stroke in which 75% of individuals after stroke experience hemiparesis of the upper and/or lower extremity. Notably, sensorimotor function is heavily reliant on proprioceptive inputs. Robotic measures estimate that 50-60% of individuals after stroke have proprioceptive deficits in position sense, perception of the body’s position, and/or kinesthesia, the perception of the body’s position during movement. Many paradigms quantify proprioception through active mirror-matching of the more affected arm; however, quantifying proprioception using only the more affected arm allows for a more direct measurement of proprioceptive acuity by avoiding the confound of interhemispheric transfer. Here, we aim to understand how proprioceptive thresholds provide information relating to motor and proprioceptive acuity after stroke, which may reveal otherwise undetected proprioceptive deficits in the stroke population. We hypothesized that proprioceptive thresholds will be predictive of motor and proprioceptive outcomes.
Individuals with stroke (N=39) and age-matched controls (N=39) completed five behavioral tasks with the KINARM robotic exoskeleton. Proprioception was assessed by 3 tasks where the more affected arm was moved by the robot. 1) Position Matching (PM) is a mirror-matching position sense assessment quantified by the variability of end point error (VAR). 2) Kinesthesia (KIN) is a mirror-matching assessment of movement speed, direction, and length quantified by initial direction error (IDE) and response latency. 3) Single Arm Proprioception (SAP) is a two-alternative forced-choice psychophysical task that evaluates the proprioceptive discrimination threshold within a single arm.
Sensorimotor function of the more affected arm, quantified by IDE, was measured by 2 tasks. 1) Visually Guided Reaching (VGR) is an 8-target center-out reaching task with vision of hand position. 2) Reaching without Vision (RwoV) is a 4-target center-out reaching task without vision of hand position. RwoV is intended to be a direct application of proprioceptive acuity in an active sensorimotor task, as the participants will have to rely on both conscious and unconscious proprioceptive feedback to understand whether the peripheral target was reached without vision of hand position. Clinical information was captured through the Functional Independence Measure (FIM), Chedoke-McMaster Stroke Assessment, and Thumb Localizer Test.
Stepwise linear regression found that proprioceptive thresholds (SAP) and VGR IDE were significant predictors of RwoV IDE (r2=0.47, p<0.001) and the FIM (r2=0.50, p<0.001). Further, proprioceptive thresholds and KIN IDE were predictors of PM VAR (r2=0.61, p<0.001). Proprioceptive thresholds have the potential to predict online proprioceptive processes that are essential for updating movements (i.e., RwoV), as well as conscious aspects of proprioception (i.e., PM). Additionally, these thresholds are useful for predicting level of functional independence as evaluated by the FIM. Overall, proprioceptive thresholds can be an indicator of sensorimotor function—as a predictor of quality of movement, proprioceptive acuity, and clinical function after stroke.
Sensory Rehabilitation
Alterations in intermuscular coordination as a potential stroke rehabilitation target using muscle synergy analysis
University of Houston, Houston, USA
Abnormal intermuscular coordination is one of the significant stroke-induced motor impairments. Muscle synergy analysis has been used to identify abnormal intermuscular coordination post-stroke. The similarity score of corresponding muscle synergies in health and stroke negatively associates with the severity of motor impairment, suggesting that targeting altered intermuscular coordination can be a potential neurorehabilitation approach. The synergy similarity score is a standard measure of alterations in muscular coordination. However, it does not indicate which specific muscular co-activations are altered in stroke, limiting to specify the target for synergy-guided rehabilitation after stroke. Thus, the study aimed to examine whether/what particular muscle co-activation within muscle synergies was related to synergy similarity, motor impairment, and quality of motor behavior measures after stroke.
We recruited twenty-four chronic stroke survivors and six neurologically intact, age-matched participants. Participants performed a 3D isometric force generation task using a paretic arm for stroke and both arms for age-matched control. We recorded end-point force and electromyographic signals of eight elbow and shoulder joint muscles. Qualities of motor behavior were evaluated using end-point force trajectories. Non-negative matrix factorization was used to identify muscle synergies. Muscle synergy similarity between stroke and age-matched control was calculated by cosine similarity. The level of co-activation of a pair of muscles within a synergy was calculated to specify the stroke-induced alteration of synergy composition. We examined a relationship between muscle synergy similarity and the level of muscle co-activation to identify one that would drive changes in the synergy similarity. The relationship between each of the two parameters and either the Fugl-Meyer Assessment (FMA) score or qualities of motor behavior was calculated by Spearman’s rank correlation to evaluate its potential as a stroke rehabilitation target.
Four muscle synergies were identified in both stroke and age-matched control, respectively. The similarity of shoulder adduction/flexion (S Add/Flex) synergy, out of the four synergies, was significantly correlated with the FMA score (p<0.01). The level of co-activation between anterior deltoid (AD) and pectoralis clavicular (PECT) muscles within the S Add/Flex synergy was strongly correlated with the similarity of S Add/Flex synergy (p<0.01). The level of AD-PECT co-activation was also significantly associated with FMA score (p<0.01) and qualities of motor behavior (p<0.05 for force tuning, number of peak speeds, duration, pathlength, and mean speed).
We identified a specific co-activation of a muscle pair that drives change in synergy similarity after stroke. Specifically, the synergy similarity decreased as the AD-PECT co-activation within an S Add/Flex synergy decreased. Also, the identified co-activation was negatively and positively associated with motor impairment and quality of motor behavior, respectively. Thus, we conclude this neuromuscular pattern underlying stroke-induced alteration in synergy composition could be considered a potential target for motor rehabilitation post-stroke.
Stroke
Plasma MicroRNA Prediction of Upper Limb Recovery Following Human Stroke
1Georgetown University, Washington, USA. 2MedStar National Rehabilitation Hospital, Washington, USA. 3. 4Spectrum Health, Grand Rapids, USA
Stroke
Intraspinal microstimulation simultaneously rebalances motor and nociceptive transmission in chronic spinal cord injury
Washington University in St. Louis, St. Louis, USA
Spinal cord injury (SCI) results in devastating physiological changes at and below the level of the lesion. For instance, SCI often leads to reduced motor output, dysregulation of spinal reflexes, impaired pelvic floor function, and SCI-related neuropathic pain (SCI-NP). Electrical spinal stimulation (ESS) therapies are a promising approach to treating these issues. However, ESS approaches for SCI recovery are generally parametrized for motor rehabilitation only. Motor-targeted ESS approaches rely on the recruitment of sensory afferent pathways and interneuron networks that provide excitatory synaptic drive to spinal motor pools. Several of these pathways are also implicated in the development and maintenance of SCI-NP. Nonetheless, these approaches often fail to assess potential off-target effects regarding spinal nociceptive transmission.
Here, we used spike train analyses to characterize the effects of motor-targeted intraspinal microstimulation (ISMS) on spinal nociceptive neural transmission after SCI. Experiments were conducted in 15 adult male Sprague-Dawley rats with chronic incomplete SCI under urethane anesthesia. All experiments were approved by the Institutional Animal Care and Usage Committee of Washington University in St. Louis. Electrophysiological recordings were conducted 6-8 weeks after a midline T8 contusion. After a T13-L2 laminectomy, microelectrode arrays were implanted at the L5 dorsal root entry zone. ISMS was delivered to the motor pools in the ventral horn, and neural transmission in sensory pathways was characterized in the superficial and deep dorsal horns. Prior to, during, and after ISMS, we mechanically stimulated the L5 dermatome by applying controlled forces ranging from innocuous to noxious. Our primary outcome measure was the change in maximum firing rate [Hz] of nociceptive specific (NS) and wide dynamic range (WDR) neuron populations (analyzed at the animal level) during nociceptive transmission. These changes were quantified by comparing the average maximum instantaneous firing rates prior to, during, and after 30min of ISMS.
We found that only one session motor-targeted ISMS in the injured spinal cord modulated neural transmission in nociceptive pathways of the dorsal horn. Indeed, ~60% of NS and WDR neuronal populations, respectively, reduced their maximum firing rate during induced nociceptive transmission after ISMS compared to before ISMS. We also found the firing rate of WDR neurons progressively decreased with increasing ISMS duration before persisting at that depressed level after ISMS was discontinued.
Our results suggest that motor-targeted ISMS retains its ability to immediately modulate spinal nociceptive transmission and to induce neural plasticity in spinal sensory networks in the chronically injured spinal cord. These actions appear to result in a net decrease in spinal responses to nociceptive sensory feedback. Although future work is required to elucidate the mechanisms underlying these actions, our results suggest that it may be possible to optimize the stimulation paradigm to deliver multi-modal therapeutic benefits after SCI.
Spinal Cord Injury (SCI)
Combined activity-based therapy and cervical spinal cord stimulation for the restoration of upper limb function after cervical spinal cord injury
1Toronto Rehabilitation Institute, Toronto, Canada. 2University of Toronto, Toronto, Canada
1. Description of the active ingredients, mechanism of actions and targets of the ABT and tCSCS, respectively, in line with the tripartite structure recommended by the rehabilitation treatment specification system framework12.
2. Delivery and tailoring of combined ABT and tCSCS, in line with the template for intervention description and replication checklist13
3. Development of detailed treatment regimen guidelines for the delivery of the ABT-tCSCS to facilitate standardization of therapy.
In tCSCS, the active ingredients were electrical stimuli. The stimulation will be delivered using MyndSearch stimulator in continuous mode at a frequency of 30-50Hz at 500-1000µs pulse width using a cathode-leading, asymmetric biphasic stimulus waveform between C3-C7. Starting with an intensity of 5mA, the current intensity will be gradually increased as per the tolerance of the patient or the best response obtained for each training task. The mechanism of action of cervical spinal cord stimulation involves stimulation of the spinal networks in the cervical region to neuromodulate the descending motor commands/motor intentions from the brain, which control the muscles. The cervical spinal cord stimulation targeted arm and hand strength, pinch, grasp, finger dexterity and prehension.
ABT and tCSCS will be delivered simultaneously. ABT-tCSCS will be delivered for 1.5 hours over 28 sessions (3xper week).
Spinal Cord Injury (SCI)
Usability of collaborative robots for rehabilitation of the upper and lower limbs after stroke and spinal cord injury: a scoping review
1Toronto Rehabilitation Institute, Toronto, Canada. 2University of Toronto, Toronto, Canada. 3University Health Network, Toronto, Canada
We searched electronic databases such as MEDLINE up to 2022. The search strategy was based on the Population-Concept-Context framework6,7 using key terms including collaborative robots, stroke, spinal cord injury, upper/lower limbs, and rehabilitation. The titles, abstracts and full-text articles identified by the search were assessed against the eligibility criteria.
The training protocols delivered by the cobots targeted range of movement, strength, and coordination for upper limb rehabilitation and ankle movements for lower limb rehabilitation after stroke and SCI. The cobots were found to be feasible, safe, reliable, and may improve upper and lower limb function after stroke and SCI.
Motor Rehabilitation
The relationship between upper extremity use at home and adherence to a home exercise program among stroke survivors
1Medical University of South Carolina, Charleston, USA. 2Ralph H. Johnson VA Health Care System, Charleston, USA
Stroke
Effects of Gait Training With and Without Electrical Stimulation on Neural, Biomechanical, and Clinical Outcomes Post-Stroke
1Georgia Institute of Technology, Atlanta, USA. 2Emory University, Atlanta, USA
Stroke
Brain functional network segregation is differentially associated with walking function in younger and older adults
University of Florida, Gainesville, USA
Motor Rehabilitation
The consideration of self-efficacy in early-stroke rehabilitation
1University of North Carolina at Chapel Hill, Chapel Hill, USA. 2UNC Health, Chapel Hill, USA
Stroke
Pairing Transcutaneous auricular vagus nerve stimulation (taVNS) and Constraint Induced Movement Therapy (CIMT) to improve motor function in infants
1Medical University of South Carolina, Charleston, USA. 2Medical University of South CarolinaRocha-Ferreira E, Hristova M. Plasticity in the Neonatal Brain following Hypoxic-Ischaemic Injury. Journal of neural transplantation & plasticity. 2016;2016:4901014-16. doi:10.1155/2016/4901014 2. Duerden EG, Foong J, Chau V, et al. Tract-Based Spatial Statistics in Preterm-Born Neonates Predicts Cognitive and Motor Outcomes at 18 Months. American journal of neuroradiology : AJNR. 2015;36(8):1565-1571. doi:10.3174/ajnr.A4312 3. O’Shea TMTM, Allred EE, Kuban KCKKCK, et al. Elevated concentrations of inflammation-related proteins in postnatal blood predict severe developmental delay at two years in extremely premature infants. The Journal of pediatrics. 2011;160(3):395-401.e4. doi:10.1016/j.jpeds.2011.08.069 4. Kimberley TJ, Pierce D, Prudente CN, et al. Vagus Nerve Stimulation Paired With Upper Limb Rehabilitation After Chronic Stroke. Stroke. Nov 2018;49(11):2789-2792. doi:10.1161/strokeaha.118.022279 5. Redgrave JN, Moore L, Oyekunle T, et al. Transcutaneous Auricular Vagus Nerve Stimulation with Concurrent Upper Limb Repetitive Task Practice for Poststroke Motor Recovery: A Pilot Study. J Stroke Cerebrovasc Dis. Jul 2018;27(7):1998-2005. doi:10.1016/j.jstrokecerebrovasdis.2018.02.056 6. Porter BA, Khodaparast N, Fayyaz T, et al. Repeatedly pairing vagus nerve stimulation with a movement reorganizes primary motor cortex. Cereb Cortex. Oct 2012;22(10):2365-74. doi:10.1093/cercor/bhr316 7. Badran BW, Jenkins DD, Cook D, et al. Transcutaneous Auricular Vagus Nerve Stimulation-Paired Rehabilitation for Oromotor Feeding Problems in Newborns: An Open-Label Pilot Study. Front Hum Neurosci. 2020;14:77. doi:10.3389/fnhum.2020.00077 8. Dawson J, Liu CY, Francisco GE, et al. Vagus nerve stimulation paired with rehabilitation for upper limb motor function after ischaemic stroke (VNS-REHAB): a randomised, blinded, pivotal, device trial. Lancet. Apr 24 2021;397(10284):1545-1553. doi:10.1016/s0140-6736(21)00475-x 9. Badran BW, Jenkins DD, DeVries WH, et al. Transcutaneous auricular vagus nerve stimulation (taVNS) for improving oromotor function in newborns. Brain Stimul. Sep-Oct 2018;11(5):1198-1200. doi:10.1016/j.brs.2018.06.009 10. Gordon AM, Hung YC, Brandao M, et al. Bimanual training and constraint-induced movement therapy in children with hemiplegic cerebral palsy: a randomized trial. Neurorehabil Neural Repair. Oct 2011;25(8):692-702. doi:10.1177/1545968311402508, Charleston, USA
All 3 infants completed the 40 hours of CIMT + taVNS intervention in 4 weeks. They tolerated the ear electrode, device set-up and stimulation well. We delivered CIMT for a mean of 114 ±8min per session, and triggered taVNS for 58% of CIMT time at a mean intensity of 0.57mA. Therapists were able to deliver CIMT reliably while also triggering taVNS stimulation with active movement during CIMT.
All 3 infants showed significant gains in assessments relative to their baseline (Fig 1). Infant 1 had marked gains in arm/hand function (QUEST Δ=40.63 vs expected Δ=5.2 with CIMT alone10) and GMFM. Infant 2 had greatest gains in gross motor function for transitional movements (GMFM Pre: 32, Post: 55, Δ=23). Infant 3 showed gains on all assessments. Gains in motor skills during CIMT+taVNS treatment were maintained at 3 months post treatment (infant 1 to date).
Motor Rehabilitation
Minimal Clinically Important Difference in Six-Minute Walk Test Distance based on Distribution Methods and Perception of a Meaningful Change in the Ease of Walking in People with Chronic Stroke
1Physical Therapy Department, University of Delaware, Newark, DE, USA. 2Biomechanics and Movement Science (BIOMS) Program, University of Delaware, Newark, DE, USA
The MCID for the 6MWT was estimated four ways, as previously used in stroke: 1) effect size,4 2) standard error of measurement (SEM),4 3) comparison of means between those who reported meaningful change in ease of walking and those who did not,4,5 and 4) receiver operating characteristic (ROC) curves with a meaningful-change GRoC score as the anchor.6 For ROC curve analysis, optimal cutoff 6MWT distance was determined using Youden’s index.6
Stroke
Short latency crossed spinal inhibition during standing in people with chronic stroke
1National Center for Adaptive Neurotechnologies/Stratton VAMC, Albany, USA. 2College of Health Professions, Medical University of South Carolina, Charleston, USA
Stroke
The Relationship Between Spatial Neglect and Balance in Adults Post-Stroke
MUSC, Charleston, USA
Stroke
Task difficulty influences paretic arm choice during goal-directed planar reaching actions after Right Hemispheric Stroke
1Moss Rehabilitation Research Institute, Thomas Jefferson University, Elkins Park, USA. 2Department of Physical Therapy, Arcadia University, Glenside, USA
Stroke
Advantages of a single motor imagery session, compared to two weeks of motor imagery training, after upper extremity peripheral nerve injury
1Washington University School of Medicine, St.Louis, USA. 2Sunnybrook Hospital, University of Toronto, Toronto, Canada
Recovery after peripheral nerve injuries in the upper extremity depends on peripheral neurophysiological factors but also requires plasticity in somatosensory and motor regions of the brain. Motor imagery (imagined movement) can induce motor learning and affect cortical sensorimotor representations, but it remains unknown whether imagery-driven cortical plasticity changes can result in improved functional recovery after peripheral nerve injury. Here we conducted a small proof-of-concept pilot study to determine whether motor imagery can improve motor performance in the affected hand of patients with unilateral upper extremity peripheral nerve injury. Patients were recruited with either acute (2-12 weeks post-surgery) or chronic (> 1 year post-surgery) injury. All participants (total n=33) practiced by imagining the movement and sensation of rapid thumb-to-finger tapping with their injured hand, alternating between 2 unique sequences (thumb to 1-4-2-3 and 4-1-3-2). Real movements were evaluated in each hand, before and after training, to determine change in finger-tapping rate from pre-training to post-training (i.e. learning). Post-training measurements also included a generalization test, measured as performance at a new finger-tapping sequence. Each practice session entailed 20 minutes of practice, including 30 blocks of 20 sec each, alternating between sequences, with 15 seconds between blocks. Two training protocols were tested: (1) a “single session” group (n = 21: 12 acute, 9 chronic) practiced for one session, with post-training measurements immediately thereafter; and (2) a “long-term” group (n=12: 4 acute, 8 chronic) practiced one session every day for 14 days, with the guidance of an online instructional video, with post-training measurements the day after their last practice session.
Our preliminary analyses indicated a main effect of training protocol on learning (single session vs. long-term, F = 9.8, p = 0.004), with no effects of patient type (acute vs. chronic, p = 0.151) or interaction (training * patient type, p = 0.443), driven by greater learning for the “single session chronic” condition compared to either “long term acute” or “long term chronic.” For generalization, we again found a main effect of single-training protocol on learning (F = 10.5, p = 0.003), with no effect of patient type (p = 0.642) or interaction (p = 0.306), driven by better generalization for the “single-session acute” condition than either “long term acute” or “long term chronic.” Overall, these results indicate that, in patients with peripheral nerve injury, a single practice session of motor sequence imagery has more impact – better learning and better ability to generalize to a new sequence – than a long course of such sessions. The apparent advantage of our single-session protocol could suggest that motor imagery practice improves performance via a short-term excitatory effect on execution, rather than via a learning process.
Peripheral Nerve/Plexus/Neuromuscular Diseases
StartReact Increases Activation of Muscles not Primarily Involved in the Task
1Department of Physical Therapy and Human Movement Sciences, Northwestern University, Chicago, USA. 2Department of Biomedical Engineering, Northwestern University, Evanston, USA. 3Interdepartmental Neuroscience, Northwestern University, Evanston, USA. 4Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago, USA
StartReact, the rapid involuntary release of a planned movement upon a startling acoustic stimulation (SAS) is likely facilitated via brainstem pathways, particularly the cortico- reticulospinal tracts (C-RSTs). While this is associated with increased muscle activation of the primary muscles involved in the movement, little is known about the effect of startReact on muscles not involved in the task. Therefore, we examined these secondary muscles in able-bodied individuals during a selective hand opening (OPEN) task and a simultaneous arm lifting and hand opening task (LIFTOPEN) task.
Due to the diffused nature of C-RSTs, we hypothesize that the secondary muscle activation will be larger in response to SAS compared to a non-startling acoustic stimulation, irrespective of whether the task engages more C-RST pathways or not.
To test this hypothesis, we collected electromyographic data from 10 able-bodied subjects (7 females, 3 males, age: 32 ± 10 years; age range: 22 – 49 years) participating in OPEN and LIFTOPEN tasks following a non-startling (80dB) warning cue and a planning phase. However, subjects were instructed to move as fast as possible if the warning cue was followed by a second sound. Startling (120dB) and non-startling (Go) sounds (80dB) were randomly given as the second sound. Biceps brachii (BIC) and flexor carpi radialis (FCR) represented the secondary muscles for both tasks. To quantify the secondary muscle activation for each muscle, the peak amplitudes within 100ms following EMG onset were located and the root mean squared value within a 50ms window around the peak was recorded as secondary muscle activation. To allow for comparison between subjects, the activation of each secondary muscle was then normalized by the EMG magnitude obtained during a maximum voluntary contraction for each participant. To test our hypothesis, a linear mixed effect model was used with motor task (OPEN vs LIFTOPEN) and sound cue (STARTLE vs GO) as predictors while BIC and FCR activations each served as outcome measures. Participants were treated as random effects.
We saw a significantly larger BIC and FCR activation in response to SAS as compared to Go, irrespective of the task (F (1, 399) = 153.5, p < 0.001 for BIC activation and F (1, 399) = 81.3, p < 0.001 for FCR activation). This large increase in BIC and FCR activation in a startReact task reaffirms the notion that startReact is facilitated via C-RSTs. Although startReact has been recognized as a promising therapeutic target to improve hand function post-stroke, the large increase in secondary flexor muscle activation in response to SAS begs the question of whether startReact can be used to improve wrist/finger extension post-stroke.
Other
Overcoming Rehabilitation Barriers During COVID-19: A Completely Virtual Tele-Exercise Intervention Study for Adults with Chronic Neurological Impairments
1Burke Neurological Institute, White Plains, USA. 2SUNY Upstate Medical University, Syracuse, USA. 3Columbia University Irving Medical Center, New York, USA. 4Weill Cornell Medicine, New York, USA
Exercise is a critical component of a healthy lifestyle, yet there are many people with chronic neurological impairments (CNI) who do not regularly exercise. Reasons include inaccessibility of gyms and equipment, transportation, and costs. COVID-19 magnified these barriers.
Participants were mailed a blood pressure (BP) monitor and Polar OH1 heart rate (HR) recording device that synched to a smartphone app. We downloaded HR data from the Polar website. Before and after each class, participants completed surveys on REDCap that queried pain level, BP, exertion, motivation, and satisfaction. Before the first session, at midpoint, after the last session, and one month later, participants completed a Reasons for Exercise Inventory, Physical Activity Log, Perceived Wellness Survey, Physical Activity Enjoyment Scale (PACES), and SF-36v2, on REDCap virtually.
The synchronous group showed greater satisfaction with use of technology in the study (90.4%) than the asynchronous group (78.4%). Adherence was higher in the synchronous group (67%) than the asynchronous group (57%). The synchronous group showed an increase in enjoyment of the exercises as the study proceeded (PACES; p=0.002), while the asynchronous group did not change in enjoyment. The synchronous group maintained their enthusiasm for exercise (REI) throughout the study, while the asynchronous group decreased in REI by the end of the study (p=0.007). There were non-significant trends toward improved health by the end of the study for both groups, indicating that a higher intensity or duration of exercise may be needed. This pilot study shows that virtual exercise and use of technology by participants are feasible in people with CNI. Synchronous classes may be preferred, because the team environment provided social interaction, immediate access to the researchers for assistance with technology, and more personalized motivation by the instructor.
Motor Rehabilitation
Beyond conjunction: Establishing spatial dissociation and association in lesion-symptom mapping
1Georgetown University, Washington, DC, USA. 2MedStar NRH, Washington, DC, USA
Lesion-symptom mapping (LSM) aims to uncover brain areas where damage reliably leads to a specific behavioral impairment. Many studies also address questions about neurocognitive architecture by finding behavioral deficits whose lesion-correlates show a spatial dissociation or colocalization (e.g. do verbal vs. nonverbal working memory rely on different substrates?). These questions often employ simple conjunction analysis, in which the intersection of two LSM results are considered on visual overlap alone. If the results overlap (conjunction), it is concluded that representations/computations are shared between the two behaviors. If the results do not overlap (disjunction), it is concluded that the representations/computations occupy distinction functional modules. However, this procedure is flawed because it does not test for differences between the maps, and presumes adequate statistical power to detect an apparent dissociation. Other approaches have involved a simple subtraction between two LSM results, or controlling for a second behavior in a first behavior’s analysis by including it as a covariate via a partial correlation. In short, many studies make claims about neural dissociations or associations but rarely have the claims been directly tested.
Here we implement a multivariate permutation-based test of the contrast between two LSM analyses, providing a rigorous procedure to establish the spatial dissociation or association of the two results. The technique is available as an update to our downloadable software package SVRLSMGUI (github.com/atdemarco/svrlsmgui/tree/svrlsmgui2). We use simulated behaviors from a cohort of left-hemisphere stroke survivors to 1) understand the different potential configurations of results that could give the impression of a spatial dissociation, and 2) probe the sensitivity to detecting dissociating results in the presence of natural lesion covariance. Specifically, three behaviors were simulated as percent damage of each individual subject’s lesion overlapping with 8mm radius spheres placed in frontal, parietal, and temporal lobes. We then tested the dissociability of each pair of simulated behaviors using our LSM dissociation procedure.
We found that an apparent spatial dissociation can occur from at least four configurations of results, each with a distinct interpretation, including a true dissociation, the case of colocalization, and two cases where antirelationships in lesion-symptom localization can incorrectly appear as dissociations. On the issue of dissociability, we gained better spatial precision when the simulation foci crossed an anatomical/vascular boundary (opposite sides of the Sylvian fissure). This finding suggests that the ability to detect distinct LSM relationships is sensitive to the sample’s lesion covariance structure, which naturally depends on a stereotyped distribution of vascular territories.
In summary, our novel method is able to detect dissociations and associations in the lesion-localization of behaviors. More fastidious investigation of the spatial configuration of lesion-symptom correlations will enhance our understanding of brain-behavior relationships and their underlying neural architecture.
Stroke
Motor Cortical Map Excitability in Persons with Chronic Traumatic Cervical Spinal Cord Injury: Relation to Maximal Volitional Activation and Upper Limb Motor Function
1Cleveland Clinic, Cleveland, USA. 2Oslo University Hospital, Oslo, Norway. 3Kessler Foundation, West Orange, USA. 4Louis Stokes Cleveland VA Medical Center, Cleveland, USA. 5MetroHealth System, Cleveland, USA
Spinal Cord Injury (SCI)
The Promise of Telerehabilitation to Increase Upper Limb Therapy Dose and Improve Continuity of Care During Early Post Stroke Recovery
1Moss Rehabilitation Research Institute, Elkins Park, USA. 2UCLA, Los Angeles, USA. 3Moss Rehab, Elkins Park, USA. 4Jefferson Health, Philadelphia, USA. 5California Rehabilitation Institute, Los Angeles, USA
Stroke
Body-Machine Interface: A Novel Virtual Robotic Platform for Controlling Assistive Devices
1Robotics Department, University of Michigan, Ann Arbor, USA. 2Physical Medicine and Rehabilitation, Michigan Medicine, Ann Arbor, USA. 3Department of Computer Science, University of Michigan, Ann Arbor, USA. 4Department of Mathematics, University of Michigan, Ann Arbor, USA. 5Department of Kinesiology, Michigan State University, Lansing, USA. 6Department of Mechanical Engineering, Michigan State University, Lansing, USA. 7Department of Biomedical Engineering, University of Michigan, Ann Arbor, USA. 8Department of Kinesiology, University of Michigan, Ann Arbor, USA
Body-machine interfaces (BoMIs) offer a robust and non-invasive alternative to brain-machine interfaces for controlling an assistive device. Compared to conventional interfaces like joysticks, BoMIs can adapt to the user’s abilities, allowing control even when the user is severely impaired. However, BoMIs typically require a prolonged learning phase for the users to understand the mapping between their body movements and machine commands. Hence, it may be a challenge for a patient to routinely visit a lab/clinic to learn to control an assistive device using a BoMI. Virtual environments for assistive devices can potentially address this issue, but there are currently limited virtual environments for high-degree of freedom assistive devices such as a commercial robotic arm, and moreover, it is unclear if the extent of learning will be similar between using a virtual and a physical device.
Here, we developed a virtual robotic platform using CoppeliaSim that replicated the functions of a commercially available physical 6-DOF robotic manipulator Jaco arm. Both the physical and virtual Jaco arm were controlled by a BoMI using signals obtained from four wireless inertial measurement units (IMUs). The IMUs were fixed to the user’s upper torso and we mapped the pitch and roll readings from each IMU into a two-dimensional (2D) velocity command. This map was determined using principal component analysis (PCA) on the IMU signals recorded while the user performed a free exploration of their torso. The robotic arm then generated the required end-effector velocity with the tip (gripper) by using dampened inverse kinematics.
Forty-three neurologically unimpaired adults practiced a target-matching task using a physical Jaco (n=25) or the virtual Jaco (n=18). Participants used their torso to move the tip of the robot to reach different targets on a circle in front of them. The experiment consisted of 3 test blocks (baseline, mid-test, and post-test), which were separated by four training blocks. We computed the average movement time (MT) and the normalized path length (PL) for each block and compared the improvements in movement performance between the two groups using a one-way ANOVA .
We found that improvements in MT (virtual: 9.9±9.5 sec and 11.1±9.1 sec; physical: 11.1±9.9 sec and 11.8±10.5 sec) and PL (virtual: 6.1±6.3 m/m and 6.6±6.2 m/m; physical: 3.3±3.5 m/m and 3.5±4.0 m/m) were similar between groups both at mid-test (MT: p=0.679 and PL: p=0.061) and post-test (MT: p=0.806 and PL: p=0.058).
Our results indicate the feasibility of using virtual environments for learning to control high-degree of freedom assistive devices. Future work is necessary to determine the extent to which virtual learning translates to the physical device and how these findings generalize to individuals with movement impairments.
Motor Rehabilitation
Characterization of ipsilateral motor evoked potentials across the chronic stroke impairment spectrum
1Cleveland Clinic Lerner Research Institute, Cleveland, USA. 2MetroHealth System, Cleveland, USA. 3Case Western Reserve University, Cleveland, USA
Motor Rehabilitation
Application of Corticomuscular Coherence in Early Stroke Rehabilitation
University of North Carolina at Chapel Hill, Chapel Hill, USA
Stroke
Effect of the upper extremity sensorimotor pathway on motor recovery and neuroplasticity with post-stroke rehabilitation
1Department of Health Sciences and Research, College of Health Professions, Medical University of South Carolina, Charleston, USA. 2Ralph H. Johnson VA Healthcare System, Charleston, USA. 3Department of Public Health Sciences, College of Medicine, MUSC, Charleston, USA. 4Department of Rehabilitation Sciences, College of Health Professions, Medical University of South Carolina, Charleston, USA
Stroke
Guided intraoperative dorsal root entry zone stimulation facilitates cortical motor evoked potentials in humans
1Columbia University, New York, USA. 2Weill Cornell Medicine, New York, USA. 3Icahn School of Medicine at Mount Sinai, New York, USA. 4James J. Peters VA Medical Center, Bronx, USA. 5Weill Cornell Medicine - New York Presbyterian, New York, USA. 6New York Presbyterian, The Och Spine Hospital, New York, USA
Spinal Cord Injury (SCI)
Alterations in Corticospinal Excitability after Stroke: A Systematic Review and Meta-Analysis
1Wayne State University, Detroit, USA. 2Michigan Medicine, Ann Arbor, USA
After a stroke, one of the underlying mechanisms theorized to contribute to motor impairment is changes in the excitability of the motor cortex, indicated by changes in motor threshold and short interval intracortical inhibition (SICI). While the motor threshold is known to increase after stroke, the evidence for the presence of increased inhibition in the affected hemisphere is somewhat conflicting. Therefore, the aim of this meta-analysis was to investigate the changes in motor threshold and SICI on the affected and unaffected sides after stroke and compare these changes to uninjured controls. A comprehensive electronic database search was performed to identify studies that objectively measured motor threshold and SICI after stroke using transcranial magnetic stimulation (TMS). Pooled standardized mean differences (SMDs) were computed using a restricted maximum likelihood meta-analysis model. The meta-analysis included data from 36 studies, with a total of 662 stroke patients and 346 healthy controls. The results showed that the motor threshold on the affected side was significantly higher than on the unaffected side (SMD = 0.94, 95% confidence interval [CI] = 0.58 – 1.31, p < 0.001) and uninjured controls (SMD = 0.96, 95% CI = 0.52 – 1.40, p < 0.001). In addition, the meta-analysis revealed a reduced SICI on both the affected (SMD = -1.52, 95% CI = -1.95 – -1.09, p < 0.001) and unaffected (SMD = -1.13, 95% CI = -1.59 – -0.66, p < 0.001) sides when compared with uninjured controls. These findings suggest that stroke leads to changes in the excitability of the motor cortex, characterized by an increase in motor threshold and a reduction in SICI. The higher motor threshold on the affected side may reflect a reduction in the excitability of the motor cortex. The reduced SICI on both sides may reflect a disinhibition of the motor cortex, which is contrary to the notion that stroke results in increased inhibition of the affected side. These alterations in excitability could pose an increased risk for abnormal movements and spasticity. In conclusion, this meta-analysis provides important insights into the changes in motor threshold and SICI after stroke and highlights the importance of assessing these parameters in the assessment and rehabilitation of stroke patients. The findings of this meta-analysis suggest that changes in motor threshold and SICI are likely to contribute to the motor impairment experienced by stroke patients and that targeting these changes may be an effective strategy for improving motor function in stroke patients
Stroke
Full-day leg movement kinematics in infants at risk of poor neurodevelopmental outcomes in rural Guatemala
1Children’s Hospital Los Angeles, Los Angeles, USA. 2University of Southern California, Los Angeles, USA. 3Wuqu’ Kawoq | Maya Health Alliance, Santiago Sacatepéquez, Guatemala. 4Brigham and Women’s Hospital, Boston, USA
Early detection of impaired growth and development in infancy is crucial for providing early interventions. Around the world, about 250 million children under five are at risk for different developmental disabilities [1], with the possible causes of malnutrition, poverty, and lack of early stimulation associated with low socio-economic status of the family. Infants from rural Guatemala (Maya peoples) are one population among them. According to reports [2], many Maya children experience chronic malnutrition, known to cause early developmental deficits that persist from infancy into adulthood. While there are interventions to address this concern in young children, none have targeted the first few months of infancy, the critical period when the impact of malnutrition on growth and development is the greatest. Currently, most diagnoses of impaired growth and development are not made until around six months of age, at which point intervention can be started.
Wearable sensors are a promising assessment method to quantitatively measure early limb movement patterns in infants at risk for poor neurodevelopmental outcomes (AR). This method is theoretically grounded as movement control is already the dominant component of neonatal behavioral assessments for the first year in healthy infants and infants with or at-risk alike for developmental disabilities. The authors have also validated the method [3]. Compared to short subjective “snapshot” assessments to identify impaired development, full-day monitoring of infants AR provides detailed, objective data on their characteristic movement control. Therefore, using this approach, this study aimed to characterize the kinematics of leg movement infants AR produced across a full day.
The study recruited 41 infants AR (F: 20). Each infant was measured three times between birth and six months of age, with one month between visits. Wearable sensors (Opals V2, APDM, Inc., Portland, OR, USA) were worn on both ankles of an infant, and a typical recording per visit lasted 10 hours across a typical day. Two kinematic variables derived from the recording were: 1) the number of leg movement per hour awake (average of left and right) and 2) peak acceleration per movement (median of the values). The leg movement count showed the median of 792 (IQR: 469). The medians of peak acceleration per movement were 2.81 m/s^2 (Left leg, IQR: 0.44) and 2.76 m/s^2 (Right leg, IQR:0.38). Nonetheless, significant age-associated increases were observed for both variables, indicating that AR in rural Guatemala generate more leg movements and higher acceleration movements as they grow. In future analyses we will determine whether early leg movement characteristics predict neurodevelopmental outcomes in these infants.
Other
Home-based Myoelectric Interface for Neurorehabilitation (MINT) conditioning to improve movement in chronic stroke survivors
1Northwestern University, Chicago, USA. 2University of Texas Southwestern, Dallas, USA
Impaired movement after stroke is due not only to weakness and spasticity, but also to abnormal muscle co-activation. We have shown that in-lab training with a myoelectric computer interface (MyoCI) that provides feedback to stroke survivors about abnormal arm muscle co-activation can enable reduction of this co-activation and associated arm impairment. The MyoCI uses surface EMGs from 2-3 muscles to control custom video games. We report here interim results from an ongoing study using a wearable version of the MyoCI, called MINT, in the home in a randomized controlled trial in the arm, as well as a new study in the leg.
Chronic stroke survivors with moderate to severe arm impairment trained for 90 min/day, 6 days/week for 6 weeks with different variants of the paradigm (2 or 3 muscles at a time) vs. a sham control group that trained on one muscle at a time. To date, 50 participants have completed training. After 6 weeks of training, participants in all experimental groups combined showed a trend of greater improvement relative to baseline on the Wolf Motor Function Test compared to sham group (mean -3.4 s vs. -1.5 s, timed portion). This improvement is better than the MCID of 1.5 s. This trend continued 4 weeks after training stopped. Participants training on 3 muscles at a time had a nearly significant improvement at 6 weeks compared to sham (-6.7 s, p=0.06, t-test) and improved significantly at 4 weeks after training stopped (-10.5 s vs -3.3 s, p=0.03). Importantly, even severely impaired stroke survivors improved with MINT conditioning.
We have begun to test MINT conditioning in the leg. Prior work by our lab suggested that hip adductor-knee extensor co-activation causes gait impairment after stroke. Four stroke survivors with moderately impaired gait used MINT in the lab to reduce abnormal co-activation between adductor magnus and rectus femoris for six 1-hour sessions over 2-3 weeks. Three more participants used MINT at home for 6 hours sessions over 1 week. Participants trained while standing with the paretic leg in near toe-off position (co-activation is highest in this position). Participants learned to isolate their muscles successfully, with a mean reduction of co-activation of 99% compared to baseline. Participants who used MINT at home experienced no safety issues. Importantly, gait speed on a 10-m walk test improved in all participants between baseline and the end of the last session. Even after this short training period, gait speed improved in all participants by 0.16 m/s, which is more than the MCID (0.1 m/s). Knee flexion angle increased and pelvic obliquity (tilt/hip hiking) decreased substantially.
These results suggest that MINT conditioning can improve arm and leg kinematics and function.
Motor Rehabilitation
Sensitrak: Automated Assessment of Forelimb Sensation in Rodents
1Columbia University, New York City, USA. 2Vulintus Inc., Lafayette, USA
Sensory Rehabilitation
Influence of motor network connectivity on walking ability in individuals post-stroke
1Ralph H. Johnson Veteran’s Affairs Medical Center, Charleston, USA. 2Department of Health Sciences and Research, College of Health Professions, Medical University of South Carolina, Charleston, USA. 3Division of Physical Therapy, College of Health Professions, Medical University of South Carolina, Charleston, USA. 4Department of Neurology, Emory University, Atlanta, USA. 5Walker Department of Mechanical Engineering, The University of Texas at Austin, Austin, USA
Abnormal co-activation patterns of normally independent muscle groups or modules during walking have been associated with decreased connectivity of the ipsilesional corticospinal tract (CST) and compensatory increased connectivity of contralesional corticoreticular pathways (CRP) in individuals post-stroke. These abnormal co-activation patterns in stroke survivors often result in fewer modules (two or three instead of the typically observed four independent muscle groups) associated with poor walking ability. Although, there is clear evidence of CST and CRP involvement in control of modules during walking in individuals post-stroke, there is limited understanding regarding the involvement of motor network connectivity.
To date 33 individuals post-stroke are included in this ongoing analysis. We collected lower extremity Fugl-Meyer Assessment, Berg Balance Scale, Functional Gait Assessment, Six-minute Walk Test, Activities-specific Balance Confidence, and self-selected walking speeds. Ground reaction force data were used to calculate Paretic propulsion (Pp) and Paretic Step Ratio (PSR). We used absolute values of Pp and PSR subtracted from 0.5 (ideal symmetry). Values further from 0 represent greater asymmetry. EMG data from 8 paretic leg muscles were collected while walking at self-selected speeds to compute the number of existing modules using non-negative matrix factorization. MRI examination for all participants included standard anatomical images and diffusion tensor imaging. Brain lesions were manually traced on the T2-weighted images. Tractography was performed using FDT’s probtrackX and streamlines connecting regions of AICHA brain atlas were computed. Previously identified regions of motor network (primary motor cortex, premotor cortex, supplementary motor area, pallidum, putamen, thalamus) were included in the analysis. The average streamlines were extracted from each pair (within and between regions of interest) for lesioned and non-lesioned hemispheres. Inter-hemispheric asymmetry was used for statistical analysis (Asymm = Lesioned/Non-lesioned). Pearson correlation test was used to identify the relationship between Asymm and the walking ability measures.
Preliminary findings show Asymm was negatively correlated with PSR (r = -0.52; p <0.01), and there was a trend for positive correlation with Activities-specific Balance Confidence (r = 0.32; p = 0.07). This suggests more symmetry of interhemispheric motor network connectivity is associated with symmetric step length and greater balance confidence. We did observe a negative association between module number and Asymm (i.e., higher Asymm values are associated with fewer modules), however, it was not significant (r = -0.20; p = 0.26). This could suggest more compensatory streamlines on the non-lesioned hemisphere are associated with fewer modules. This would be consistent with our previous findings demonstrating that individuals with two modules rely on their contralesional CRP following damage to ipsilesional CST. Continued enrollment to reach our desired sample size and including functional MRI is warranted to have sufficient statistical power and better understand the control of modules in individuals post-stroke with different impairment levels.
Stroke
The Effectiveness of Temporary Deafferentation for Upper Limb Rehabilitation in a Patient with Spinal Cord Injury: A Case Study
The University of Texas Rio Grande Valley, Edinburg, USA
Spinal Cord Injury (SCI)
Spinal motor neuron characteristics & disease progression in ALS: a lower limb focused descriptive study
1University of Illinois at Chicago, Chicago, USA. 2The University of Chicago Biological Sciences, Chicago, USA
Peripheral Nerve/Plexus/Neuromuscular Diseases
Effects of repeated exposure to novel gait perturbations on post-stroke walking balance
1Medical University of South Carolina, Charleston, USA. 2Ralph H. Johnson VAMC, Charleston, USA
Outcome measures were assessed prior to beginning the intervention, every 4-weeks during the intervention, and in a follow-up session 12-weeks after the intervention. We quantified paretic foot placement modulation by calculating the partial correlation between mediolateral foot placement location at the end of a step and mediolateral pelvis displacement at the start of the step, accounting for pelvis velocity. We assessed walking balance with the Functional Gait Assessment (FGA) and balance confidence with the Activities-specific Balance Confidence scale (ABC). Changes in outcome measures were assessed within each group with repeated measures ANOVAs, using data from individuals who completed all assessment sessions (n=44).
Changes in clinical assessments were more equivocal. The Perturbing group exhibited significant increases during the intervention in both FGA (p<0.0001) and ABC (p=0.006). The Control group exhibited a significant increase in FGA (p=0.023) but not ABC (p=0.77). The Assistive group did not exhibit significant changes in either FGA (p=0.081) or ABC (p=0.07).
Motor Rehabilitation
Left/right hand choices are driven by a combination of motor and non-motor difficulty
Washington University School of Medicine, Saint Louis, USA
Reach-to-grasp actions are fundamental to the daily activities of human life. However, few methods exist to assess individuals’ reaching and grasping actions in unconstrained environments. The Block Building Task provides an opportunity to directly observe and quantify the reaching to grasping actions as well as left/right hand choices. The Block Building Task has previously been used in healthy adults to identify the continuous nature of hand preferences (e.g. Stone et al. 2013), and in peripheral nerve injury patients to illustrate the stability of hand preferences despite injuries to the dominant hand (Philip et al. 2021).
Here, we created and tested new variants of Block Building Task to identify features that can serve as a “task difficulty” axis to modulate hand usage for reach-to-grasp action, and/or hand usage’s relationship with motor performance. Forty-four healthy adults (age 25.9 ± 6.7) participated in this adequately-powered cross-sectional single-arm study. Participants sat at a table with 40 Lego blocks. Participants were presented with a model and were asked to build the model as quickly and accurately as possible, with no cues/feedback about the task. Each participant completed 4 variants of the Block Building Task in a 2x2 design that varied Block Size (Normal, 2.2 ± 1.4 cm3; vs. Small, 0.8 ± 1.6 cm3) and Quantity (10 vs. 5 blocks per model). Variants were presented in counterbalanced order, with consistent block locations across participants. To measure motor performance, participants completed a precision drawing iPad app (Philip et al. 2023) with each hand.
As expected, our 4 Block Building Task variants led to different patterns of left/right hand choices. The main effect of Model Size was not significant (p = 0.365). However, the main effect of Block Size was significant (p = 0.013), as was the interaction (p = 0.002). Post-hoc analyses of the interaction indicated that, for complex models (Quantity = 10), participants used their dominant hand more for Small blocks than Normal blocks, but for simpler models (Quantity = 5), Block Size did not significantly impact hand choice. That means that smaller blocks forced participants to rely more on their dominant hand, but mostly when models were more complex.
Hand choice and precision drawing performance did not correlate for any measure of drawing performance (r2 < 0.07), confirming previous findings that hand usage is largely independent of performance (Philip et al. 2021).
Together, these results show that left/right hand choices are driven by a combination of motor difficulty factors (Block Size) and non-motor difficulty factors (Quantity, which may represent visual or working-memory complexity). While reach-to-grasp hand choices are primarily independent of motor performance in healthy right-handed adults, a task with multi-faceted difficulty can drive individuals to rely more on their dominant hand.
Motor Rehabilitation
Method for Training Assessors and Maintaining Reliability for Upper Extremity Clinical Assessments
1Department of Health Sciences and Research, College of Health Professions, Medical University of South Carolina, Charleston, SC, USA. 2Ralph H. Johnson VA Healthcare System, Charleston, SC, USA. 3Department of Public Health Sciences, College of Medicine, Medical University of South Carolina, Charleston, SC, USA. 4Department of Rehabilitation Sciences, College of Health Professions, Medical University of South Carolina, Charleston, SC, USA
During clinical assessments for study participants, each assessment is videoed and scored live by the assessment administrator. Videos are later coded to ensure raters remain blinded to both group assignment and assessment timing before or after intervention. Coded videos are distributed to blinded raters for scoring. Scores are compared between raters. If scores do not align with WMFT time <1 sec and ARAT score <1, the trainer reviews the video and asks raters to rescore the video with a reminder for specific scoring rules. If rescores continue to differ, the trainer meets with the rater to clarify interpretation of the video or scoring criteria and add this clarification to the training protocol as needed. A final score is determined by the rater during this meeting.
Stroke
Relationship between Activity-based Corticocortical Connectivity and Upper Limb Motor Function in Stroke Survivors
1Medical University of South Carolina, Charleston, USA. 2Claflin University, Orangeburg, USA. 3Ralph H. Johnson VA Healthcare System, Charleston, USA
Stroke
The contributions of executive function to automaticity and attention allocation during dual tasking in individuals with Parkinson’s disease
Saint Louis University, Saint Louis, USA
Other
Noninvasive vagus nerve stimulation (taVNS) increases feeding volumes and white matter micro structure in infants slated for G-tube
1Medical University of South Carolina, Charleston, USA. 2King Saud bin Abdulaziz University for Health Sciences, KSA, Jeddah, Saudi Arabia
Results: Mean GA at taVNS start was 43±3.8 weeks. Mean number of days trying po prior to taVNS was 46±18d. Overall, daily po volumes were significant increased during compared with prior to taVNS, and 54% (19/35) of infants achieved full oral feeds (responders). Days to full po feeds were significant shorter 9.6±7.8d (2X) vs 15.5±7.8d (1X, p<0.05] and increase in po volumes significantly greater [F 8.05, p = 0.01] with twice daily treatment. Diffusion kurtosis in corticospinal tract at the cerebellar peduncles shows significant treatment response and predicts attaining full oral feeds.
Motor Rehabilitation
Motor and cognitive deficits reduce the ability to modulate spatiotemporal aspects of gait in individuals with mild cognitive impairment
1Emory University, Atlanta, USA. 2Atlanta VA Center for Visual & Neurocognitive Rehabilitation, Atlanta, USA. 3Birmingham/Atlanta VA Geriatric Research Education and Clinical Center, Atlanta, USA
We quantified movement accuracy for each trial as the average percent error across spatial (peak hip, knee, and/or ankle angles) and/or temporal (step timing) variables relative to trial-specific target values (e.g., 90° peak swing-leg hip flexion). Percent error was averaged across trials within each modification class (spatial, temporal, spatiotemporal). Separately for each class, we tested for group differences in movement accuracy (vs. HOA) using fixed-effects linear models with indicator variables denoting the HYA and MCI groups. We report group differences vs. HOA as average percent errors (Δ).
Motor Rehabilitation
Characterization of changes to inter-joint active and passive couplings in the arm and hand following stroke
1Department of Bioengineering, University of Maryland, College Park, USA. 2Department of Physical Therapy & Rehabilitation Science, University of Maryland, Baltimore, USA. 3University of Maryland Rehabilitation and Orthopaedic Institute, Baltimore, USA. 4Department of Orthopaedic Surgery, Baltimore, USA
Over the last decade exoskeletal rehabilitation robots have become ever more popular. Robotic therapy has established itself as a new and powerful tool for stroke rehabilitation, as it can provide a deeper understanding of the complex neuromechanical properties of the patient’s limb movement control, as well it can make the rehabilitation more effective and efficient. We have developed an arm exoskeleton capable of actuating the shoulder, elbow, wrist and metacarpophalangeal (MCP) joint and simultaneously measuring the torques and forces at the multiple joints.
Stroke
Associations Between Posterior Parietal and Motor Cortical Thickness and Obstacle Negotiation in Older Adults
1Malcom Randall VA Medical Center, Gainesville, USA. 2University of Florida, Gainesville, USA
Motor Rehabilitation
Feasibility of Interleaved Computerized Cognitive Training and Accelerated, High-Dose Repetitive Transcranial Magnetic Stimulation in Amnestic Mild Cognitive Impairment
Medical University of South Carolina, Charleston, USA
Results: All participants completed BrainHQ exercises during the acute iTBS-rTMS treatment portion of the study and most continued to engage in the exercises during the follow-up period. During the treatment phase, participants completed a median of 23 minutes of BrainHQ on treatment day 1, 38.5 minutes on treatment day 2, and 42.8 minutes on treatment day 3. Of those who engaged in BrainHQ over the 4-week follow-up period, the amount of time was variable, ranging from a total of 5.4 – 392.7 minutes per week. Interestingly, we observed increases over the follow-up period in both the number of participants engaging in the exercises (66.6% in week 1, 85.7% in week 2, 81.0% in week 3, 90.5% in week 4) and the median number of minutes (34.3 minutes during week 1 to 124.9 minutes during week 4).
Cognitive/Language Rehabilitation
Score Card for Reporting Individual Performance Post Stroke
1MUSC, Charleston, USA. 2Ralph H. Johnson VA Medical Center, Charleston, USA
Individuals with stroke often ask about their progress when participating in rehabilitation research studies. Providing relevant information about their function can increase both participant knowledge and satisfaction. However, sharing research results with participants is often limited by several barriers: the complex nature of research outcome measures; the long time periods needed to assess the effects of an intervention; the risk of misrepresented or misunderstood clinical meaningfulness, due in part to limited health literacy; and the risk of participants expecting clinical advice that goes beyond the scope of the research study. The objective of this project is to develop a Research Report Card through which various approaches to providing participants with information can be tested, providing feedback about participant performance on a variety of standardized assessment outcomes.
As part of an ongoing study, individuals who have experienced a stroke complete a study visit once per year for five years, in which we use standardized assessment outcomes to measure various aspects of function that can be affected after stroke including balance (Berg Balance Scale; BBS), mood (Patient Health Questionnaire 8; PHQ-8), memory (Montreal Cognitive Assessment; MoCA), and reaching (Upper Extremity Fugl-Meyer; UEFM). Thus far, 135 participants have been enrolled.
The enrolled participants exhibited a high level of variability in their level of function across the assessed outcome measures. Specifically, the median was 38.5 for the BBS (min 5, max 56), 10.5 for the PHQ-8 (min 0, max 23), 19.5 for the MoCA (min 9, max 29), and 28.5 for the UEFM (min 0, max 60 max). Based on the collected information, we have developed two strategies for providing participants with a personalized Research Report Card that combines an easily understood description of the assessment with participants’ individual assessment scores and available population-wide information. The first strategy graphically depicts individual assessment scores, along with the context of Minimal Detectable Change values and Minimal Clinically Important Difference scores. The second strategy graphically depicts individual scores, along with established population categories (e.g., balance scores that indicate a fall risk).
Our development of a Research Report Card for individuals with stroke is just a first step toward our goal of providing research participants with useful information. The design is intended to provide clarity and context to assessments that participants are unlikely to be familiar with, and to allow changes over time to be visualized. A necessary next step is to gather feedback on this design from stakeholders, including rehabilitation researchers, clinicians, and research participants. Through iterative improvement, our goal is to create a feedback template that can be integrated into a wide range of research studies.
Stroke
Visuospatial cognition predicts performance on a complex obstacle walking task in older adults
1Malcom Randall VA Medical Center, Gainesville, USA. 2University of Florida, Gainesville, USA. 3Drexel University, Philadelphia, USA
Walking performance is correlated with cognitive function in older adults, and performance on both task types declines with age. Walking is a dynamic task which requires various levels of neural resources for safe, effective, and autonomous ambulation. Research has shown associations between typical, overground walking as well as complex walking task and cognitive function. However, measuring the cognitive control of walking is challenging for researchers. Currently, the most popular way to determine cognitive resources utilized during walking is with dual-task walking, in which walking is combined with a secondary task (e.g. counting backwards by sevens). While dual-task assessments have been incorporated widely this paradigm has shortcomings, including scaling the task difficulty to mitigate a ceiling or floor effect and controlling for task prioritization. The objective of the current study was to design a novel assessment in which difficulty can be easily manipulated. Here we present the Obstructed Vision Obstacle (OBVIO) task, which asks participants to walk through a 10-meter obstacle course while holding a lightweight posterboard such that the board obstructs their vision of approaching obstacles (similar to walking with a tray). In obstructing the foam obstacles we are intentionally having participants rely on cognitive resources to remember where the foam blocks are in order to avoid striking them. This study included 71 older adults (55 male) with an average age of 76 +/- 7 years. The average preferred 10-meter walking speed was 1.19 +/- 0.33 m/s with a fastest safe walking speed average of 1.56 +/- 0.32 m/s. We report that use of the obstructed vision board significantly slowed participants down an average of 21% (0.22 m/s, p<0.001) and they hit 177% more of the obstacles (p<0.001). OBVIO walking performance (a score based on both speed and accuracy) was significantly correlated with a computer-based test of visuospatial working memory (r=.27, p<0.05), as well as tests of visual attention (r=0.42, p<0.001) and verbal working memory (r=0.25, p<0.05). Additionally, these three domains predicted the hindrance of the obstructed vision board, which is how much walking performance decreased with addition of the tray (vsWM: r=0.25 p<0.05, VA: r=0.27 p<0.05, vWM: r=0.24 p<0.05). These findings provide initial support for OBVIO as an effective task to test older adults’ cognitive control of walking.
Motor Rehabilitation
Motor-sensory network correlates for lower extremity impairment and gait speed in chronic stroke
1King’s College London, London, United Kingdom. 2VA Northeast Ohio Health System, Cleveland, USA. 3Case Western Reserve University School of Medicine, Cleveland, USA. 4University Hospitals of Cleveland, Cleveland, USA
Stroke
Estimating compensatory truncal movements in healthy controls and patients withweakness due to recent stroke using gyroscope data from wearable sensors
1University of Rochester, Rochester, USA. 2University of Rochester,, Rochester, USA. 3Washington University School of Medicine, St. Louis, USA. 4Department of Neurology, University of Rochester, Rochester, USA. 5Department of Physical Therapy, Nazareth College, Rochester, USA
First, data was collected from healthy control subjects performing UE exercises while in a movement analysis research laboratory. Gyroscope data during specific parts of the exercises was compared with truncal movement as measured by video motion capture technology. Then, gyroscope data was analyzed from wearable sensor data collected from subjects with UE weakness due to recent (<4 weeks) unilateral stroke performing the same exercises.
Motor Rehabilitation
Upper Extremity Movement Smoothness Maps onto Motor Function and Injury after Acute Stroke
1Harvard University, Cambridge, USA. 2VA Medical Center, Providence, USA. 3Massachusetts General Hospital, Boston, USA. 4Brown University, Providence, USA. 5Northeastern University, Boston, USA
Upper extremity (UE) motor impairments are a common and major source of functional disability for stroke survivors. Current outcome measures for assessing motor function do not capture trial-by-trial variability or within session effects. There is a need for high-resolution biomarkers that capture variability in UE movement after stroke. Goal-directed UE movements after stroke exhibit spatial and temporal discontinuities (i.e., abnormalities in smoothness). The aim of this study was to examine movement smoothness as a biomarker of UE movement after acute stroke, when motor impairment is most severe and behavioral impairments most directly reflect neural injury. We related 2D movement smoothness to UE motor impairment and other functional assessments as well as to patterns of neuroanatomic injury after acute stroke.
Twenty-three participants with UE motor impairment after acute stroke (4.6 ± 1.9 days poststroke, Fugl-Meyer range [4-65]) completed 80 trials of a planar reaching center-out task (8-directions) with their more affected UE on the Bionik InMotion2 Arm Therapy System, a clinical rehabilitation robotic system. For comparison, twenty-three able-bodied adults completed the same planar reaching task. Movement smoothness of planar reaching trials was quantified using Spectral Arc Length (SPARC), a well-validated measure. Differences in smoothness between able-bodied and acute stroke trials were analyzed using a Wilcoxon Rank Sum Test. We characterized trial-by-trial variability in smoothness with (1) standard deviation and (2) the percent of trials that fell within 1.48*IQR of the able-bodied median smoothness. We investigated relationships between smoothness (mean and variability) and UE outcome measures (Fugl-Meyer, Elbow Extension/Flexion Strength, Modified Rankin Scale, Barthel Index, Box and Blocks and 9-Hole Peg). Finally, we used voxel-based lesion symptom mapping (VLSM) to relate smoothness to patterns of neuroanatomical injury.
Acute stroke and able-bodied control groups were found significantly different in smoothness during planar reaching (z=-22.98, p<0.001). We found significant associations between UE clinical outcome scores and average smoothness (correlation estimates range from 0.46 to 0.80, median=0.57). There was a substantial amount of trial-by-trial variability in smoothness (standard deviation values range from 0.09 to 1.63, median=0.34). Generally, patients with moderate-severe motor impairment had the most smoothness variability and, on average, 61% of their trials fell outside of the able-bodied range. Disrupted movement smoothness was associated with injury to critical areas of the motor system, specifically the primary motor cortex and the posterior limb of the internal capsule.
In conclusion, individuals with acute stroke had abnormal movement smoothness: reduced and more variable smoothness was associated with more impaired UE motor function. Disrupted movement smoothness mapped onto the motor system, particularly regions with dense motor fibers. Taken together, these findings support movement smoothness as a robust, real-time biomarker of UE motor function after stroke.
Stroke
Neurophysiological Effects of Trigger Point Deep Dry Needling of Latent Trigger Points
1The Medical University of South Carolina, Charleston, USA. 2University of St. Augustine, Miami, USA. 3Universidad Rey Juan Carlos Facultad de Ciencias de la Salud, Madrid, Spain
Other
Effects of contralesional motor cortex LF-rTMS on learning a skilled hand task in the subacute phase post stroke
1Emory University, Atlanta, USA. 2University of Georgia, Athens, USA. 3Augusta University, Augusta, USA
Primary motor cortex in the hemisphere contralateral to the stroke (contralesional M1) may serve as a source of cortical reorganization and related recovery1-5 and has been targeted by non-invasive brain stimulation in rehabilitation treatment6. For example, low frequency repetitive transcranial magnetic stimulation (LF-rTMS) seems to have a beneficial effect on affected upper extremity (UE) function when applied prior to physical therapy7 but results have been mixed8 and LF-rTMS has been shown to disrupt learning in healthy participants9. Because motor learning is a major mechanism underlying training-related improvement in patients after motor stroke5, a better understanding of the effects of LF-rTMS on learning post-stroke is important.
In this randomized double-blind sham-controlled study of participants with a single cerebral ischemic stroke 4 +/- 2 weeks affecting M1 or its output system and corresponding hand paresis (N = 17, 10F, 58.4 ± 9.6 years) and healthy right-handed age-matched controls (N = 20, 13F, 60.1 ± 7.2 years), the effect of LF-rTMS on learning a skilled hand motor task was determined.
As previously described10-12, participants learned to manipulate a joystick to move a cursor into a target with each hand. Target size varied to allow for parametric variation in demand on precision. Two runs of 252 trials for each hand. were completed. Between the two runs, 15 minutes of 1 Hz rTMS at an intensity of 90% motor threshold or sham stimulation was applied to the ECU hotspot of left M1 (LM1) in healthy participants or contralesional M1 in participants after stroke. Session order (sham/rTMS) was counterbalanced across participants. Mean pointing task accuracy was determined for each run separately.
A 2 (Group: healthy/stroke) x 2 (Time: pre/post) x 2 (Hand: affected/left or nonaffected/right) x 2 (Condition: rTMS/sham) repeated measures ANOVA was performed to analyze how LF-rTMS affected accuracy and motor learning. We found significant main effects of Group, Hand, and Time and a significant interaction between Group, Time, and Condition. Accuracy was higher in the healthy than stroke group, higher in the non-affected/right hand than the affected/left hand, and higher post LF-TMS than pre LF-TMS. Further investigation of the interaction revealed that the healthy group showed that learning was preserved across stimulation with sham but disrupted with LF- rTMS, while the stroke group showed preserved learning in both conditions.
These findings confirm for healthy participants the involvement of LM1 in the early phase of learning and its disruption by LF- rTMS when applied after skill learning without affecting subsequent learning or performance9. In contrast, LF- rTMS applied to contralesional M1 of participants with stroke has no effect on learning or performance. This is consistent with the line of evidence of abnormal contralesional M1 activity11, 13 and needs additional study.
Stroke
Potential Mechanisms of Stiff-Knee Gait in Individuals Post-stroke: A Narrative Review
1Department of Physical Medicine and Rehabilitation, Case Western Reserve University, Cleveland, USA. 2MetroHealth Center for Rehabilitation Research, MetroHealth Hospital, Cleveland, USA
Stiff-Knee gait (SKG) is a gait dysfunction commonly observed post-stroke that is characterized by decreased swing phase knee flexion angle. Importantly, SKG can impair walking, increasing fall risk and metabolic cost reducing overall activity/participation. The current standard pharmacological interventions for post-stroke SKG aim to reduce quadriceps spasticity that is supposedly preventing knee flexion. While both botulinum toxin and baclofen have demonstrated positive outcomes for clinical measures of spasticity, studies have observed only a limited increase in knee flexion and conflicting results for other functional gait outcomes. These mixed results combined with recent observational evidence suggests that there are other potential causes requiring the need to re-examine the etiology of SKG. The aim of this review is to compile and appraise the current state of the science regarding the neurophysiology and biomechanics of individuals post-stroke with SKG. To this end, the review is organized in the following manner to achieve specific objectives towards the understanding of SKG: 1) operationally define SKG and further detail its presentation which lacks appropriate specificity for such heterogenous presentations, 2) describe the neurophysiological components that relate to the biomechanics of SKG including evidence supporting proximal and distal contributors to SKG, and 3) highlight future directions necessary to the elucidation of SKG etiology and hope to inspire innovations in personalized rehabilitative interventions.
Stroke
Understanding the mechanisms of action observation as a rehabilitation intervention for stroke
1Georgia Institute of Technology, Atlanta, USA. 2Georgia State University, Atlanta, USA
Someone in the United States has a stroke every 40 seconds. Almost 70% of people who have a stroke experience the loss of arm and hand movement which ultimately decreases daily function and contributes to poor quality of life. Thus, there is a great need for rehabilitation in the stroke population. Action observation (AO) has emerged as a potentially powerful therapeutic tool to improve stroke rehabilitation by allowing a stroke survivor to learn how to use their arm and hand again by watching and attempting to replicate “normal” arm and hand movements. AO training in patients with upper extremity amputations revealed that their movement is improved when they focus their eye gaze on specific aspects of the movement in the video that may maximally benefit the patient. It is unknown if these eye gaze patterns may serve as a mechanism behind AO and how stroke severity impacts these gaze patterns. In this study, we seek to identify whether there is a relationship between gaze patterns and motor behaviour during AO in stroke survivors with varying levels of impairment. We recruited 20 stroke survivors with upper limb impairment into the study with upper extremity Fugl-Meyer assessment scores ranging from 23 to 54. Participants completed repeated trials of action observation followed by action execution of picking up small, round discs one at a time, transporting the discs across and above a barrier, and placing the discs correctly in the designated coloured, circular slot. We hypothesized that AO augments visuomotor strategies to help support improved performance (or decreased errors) and that there are different gaze strategies over time and across the the less affected and more affect upper extremities. Preliminary analysis revealed improvement in task performance over time with repeated trials, on both the more and less affected sides. In addition, task performance improved with increased shoulder movement in the horizontal and vertical plane, suggesting kinematic adaptations to produce a better performance score. During action observation of the video, gaze data revealed distinct points of interest in the video depending on the severity of the individual and the side performing the action. On the more affected side, the mildly impaired participants focused gaze more on the arm of the trainer, while the moderately impaired participants focused gaze more on the performer’s hand and the act of picking up the disc. For the less affected side, there are more similarities across stroke severity where most participants will look at the starting point during action observation while looking at the end goal more during action execution. This suggests that AO impacts the perception of observed movement based on severity, which may contribute to improvements in the methods of AO administration to foster increased movement.
Stroke
Multi-Joint Assessment of Arm Proprioception Impairments Post Stroke
11Department of Physical Therapy & Rehabilitation Science, University of Maryland, Baltimore, USA. 2Department of Bioengineering, University of Maryland, College Park, USA. 3University of Maryland Rehabilitation and Orthopaedic Institute, Baltimore, USA. 4Department of Physical Therapy & Rehabilitation Science, University of Maryland, Baltimore, USA. 5Department of Orthopaedic Surgery, University of Maryland, Baltimore, USA
Somatosensory impairment impacts negatively on motor recovery and movement control in stroke survivors. Previous studies on touch and proprioception focused on characteristics of an individual joint1 but there has been a lack of characterization of multi-joint proprioception. Daily activities of human motion involve multiple joints simultaneously in the upper limb, therefore, examination and treatment of multiple joints are more critical. Objectives of this study was to examine proprioceptive impairments at the shoulder, elbow, and wrist joints of stroke survivors using a multi-joint robot device in order to better understand multi-joint sensory deficiency and facilitate sensory-motor recovery post-stroke. Methods: sixty-nine stroke survivors: 54.3±13.9 (mean±SD) years, and 41 healthy controls; 47.4±16.2 years participated in this study. Inclusion criteria of the stroke survivors were diagnosis of a single stroke (hemiplegic and post-stroke duration was within one year with cognitive ability to follow simple instructions). The impaired arm of the stroke survivors and the dominant arm of the healthy controls were tested in this study. All participants signed informed consent.
A multi-joint robotic arm was employed for evaluating proprioception acuity of the shoulder horizontal adduction/abduction, elbow and wrist flexion/extension. Each participant was seated with the initial position of 70 deg shoulder horizontal adduction, 60 deg elbow flexion, and 0 deg wrist flexion. One of the three joints was randomly selected and moved slowly at 0.5 deg/s inward or outward, with a total of 18 trials (3 joints, 2 directions/joint, 3 repetitions/condition). The participant was asked to press a handheld switch as soon as he/she felt movement in a joint and report which joint and in which direction it had moved. Proprioceptive acuity was measured using the threshold detection of passive motion.
Stroke
Heteronymous spinal pathways between quadriceps and soleus in stroke survivors. A comparison between nerve and muscle stimulation
1Division of Physical Therapy, Department of Rehabilitation Medicine, Emory University, Atlanta, USA. 2Exercise and Rehabilitation Sciences Laboratory, School of Physical Therapy, Faculty of Rehabilitation Sciences, Universidad Andres Bello, Santiago, Chile. 3Center for Visual and Neurocognitive Rehabilitation, Atlanta VA, Atlanta, USA
Abnormal joint coupling arising from muscles active together at the wrong time is a common impairment after stroke. A possible explanation for synergistic muscle activations is altered heteronymous spinal reflex circuits, which normally provide excitatory and inhibitory feedback between muscles. For example, femoral nerve stimulation (FNS) normally produces short latency excitatory followed by inhibitory feedback onto soleus in healthy adults. Recently, Dyer et al. found markedly increased excitation with minimal inhibition in persons after stroke which could contribute to excitatory coupling of knee and ankle extensors. However, the reason for the marked heteronymous excitation (i.e., larger excitation and/or reduced inhibition) cannot be determined using FNS alone, because nerve stimulation activates excitatory muscle spindle afferents and inhibitory pathways. Based on our recent work showing that quadriceps muscle stimulation can evoke heteronymous inhibition with minimal excitation onto SOL in healthy young adults, this study aims to determine the fate of heteronymous inhibitory feedback after stroke by comparing the effects of femoral nerve and quadriceps muscle stimulation (QS) onto ongoing SOL EMG in stroke survivors and matched controls. We hypothesized a greater magnitude and frequency of heteronymous excitation onto SOL from FNS compared to QS, and larger excitation on the paretic compared to non-paretic limb and to healthy control limbs. We also hypothesized that heteronymous inhibition onto SOL will be greater with QS compared to FNS and largest in healthy control limbs and non-paretic limbs compared to the paretic limbs. Participants sat wearing an ankle immobilizer boot. Heteronymous feedback was examined by stimulating FN or Q in separate trials while participants held 20% SOL MVIC. FNS intensity was 2x motor threshold and QS intensities were selected to match knee extension torque between stimulation conditions. Heteronymous feedback was examined bilaterally in persons with stroke and on one side for controls. Preliminary results indicate heteronymous excitation was more frequent in the paretic compared to non-paretic and with FNS compared to QS. Excitatory magnitudes were not larger on the paretic limb compared to the non-paretic. In contrast to our hypothesis, preliminary data indicate no differences in heteronymous inhibition between limbs in stroke survivors or with controls. Larger inhibition was consistently elicited with FNS compared to QS. A better understanding of heteronymous circuitry may help clarify their functional role during movement and their role after stroke.
Stroke
Cortical Map Representation of the Motor Evoked Potential and Silent Period for the Ankle Dorsiflexor Tibialis Anterior in People With and Without Chronic Incomplete Spinal Cord Injury
Medical University of South Carolina, Charleston, USA
Motor evoked potential (MEP) to transcranial magnetic stimulation (TMS) is frequently used as a measure of corticospinal excitability and plasticity, which are essential in motor skill learning and re-learning after CNS injury. While MEP uses increase, its relation to the post-MEP silent period (SP) that reflect corticospinal inhibition in cortical representation remains largely unknown. Thus, as the first step towards understanding cortical representation of corticospinal excitation and inhibition in health and disease, we examined cortical MEP and SP maps for ankle dorsiflexor tibialis anterior (TA) in adults with foot drop due to chronic incomplete spinal cord injury (SCI) and without known neurological conditions.
Nine adults (56±6 yrs) with chronic (1-15 yrs post) incomplete SCI and 12 adults (23-30 yrs) without neurological conditions participated in this study. EMG electrodes were placed over the TA and soleus of the more affected leg (SCI) or the right leg (non-SCI). MEPs were elicited using Magstim 200-2 and a custom-made bat-wing coil with radii of 9 cm, while the sitting participant maintained TA EMG activity at the preset level (≈15% maximum isometric voluntary contraction [MVC] level for non-SCI and ≈30%MVC for SCI) with ankle, knee, and hip joints fixed at ≈-10º, 60º, and 70º, respectively. Using the TMS intensity of 10-15% above threshold estimated at a tentative optimum location, TA MEP and SP were mapped over ±3cm posterior to the vertex and -1 to 4cm contralateral to the studied leg. Four stimuli were applied at each of the 42 locations. MEP was measured in peak-to-peak amplitude over TA MEP window (e.g., 30–50ms post TMS in non-SCI).
SP was measured as the duration between the time of TMS and the return of EMG activity (from post MEP suppression) to the peristimulus level.
MEP hotspot was commonly found at 1.5cm lateral and 1cm posterior to the vertex in non-SCI and 2cm lateral and 1cm posterior to the vertex in SCI; SP hotspot differed from MEP hotspot in no particular direction by 1.7 cm on average in non-SCI and by 1.8 cm in SCI. When we calculated map areas over which the measured responses were >50% peak response, MEP map area was significantly larger in SCI (75±14% of the total mapped area of 30 cm2) than in non-SCI (38±9%) (p<0.0001 by t-test). In SCI, SP map area (59±18%) was smaller than MEP map area (p<0.05) and not different from non-SCI SP map area (46±18%, p= 0.12).
Disproportionately broader cortical representation of corticospinal excitation than inhibition in SCI may reflect reduced effectiveness of excitatory corticospinal drive in this population. Further analyses are currently ongoing to better understand the balance and relationship between corticospinal excitation and inhibition in people after SCI.
Spinal Cord Injury (SCI)
More than Meets the Eye: Calibrating Computer Vision for Post-Stroke Upper Limb Movement
University of Kentucky, Lexington, USA
Approximately 795,000 people experience a stroke in the US annually [1]. An estimated 85% of these individuals have subsequent difficulty moving an upper limb, and this leads to known reductions in quality of life [2,3]. In stroke rehabilitation research, measuring quality of movement (QoM) of the upper limb is a recommended practice [4]. Computer vision is a promising technology to measure QoM as this technology eliminates the need for any instruments placed on the person [5]. However, problems arise during the calibration process of computer vision. This calibration process is a critical step for computer vision to extract information from an image—to see more than meets the eye. In this case study, we present an application of computer vision to obtain post-stroke upper limb QoM metrics, and we highlight an instance of calibration failure. We then describe a subsequent technique to remedy this calibration failure. Finally, we apply this technique to demonstrate the potential of computer vision to measure upper limb QoM in a neurotypical participant.
Stroke
Evidence-based infant assessment for cerebral palsy: relationship to early diagnosis and intervention access
1University of Minnesota Twin Cities, Minneapolis, USA. 2University of Wisconsin-Madison, Madison, USA
Diagnosis of cerebral palsy (CP) has historically been delayed until children are around 2 years old, potentially missing a period of heightened neuroplasticity to influence lifelong outcomes. Recent clinical guidelines (2017) highlight the importance of key evidence-based assessments for early detection of CP, with the possibility of providing a diagnosis under 6 months of age. Magnetic Resonance Imaging (MRI) can identify injury patterns associated with increased risk of CP. The Hammersmith Infant Neurological Examination (HINE) and the General Movements Assessment (GMA) provide insight into neuromotor development in early infancy. A combination of these assessments is likely to provide greater accuracy in diagnosis.
The objective of this study was to determine the relationships between early evidence-based assessment, timeline of diagnosis, and access to intervention in a population of infants at high-risk for CP. Retrospective chart review and analysis was performed with patients seen between 2010 and 2022 at the University of Wisconsin Newborn Follow Up Clinic, an interdisciplinary outpatient clinic in Madison, WI specializing in developmental assessment in the first three years of life. Chart review was performed for children who received a CP diagnosis from a developmental pediatrician to assess: 1) age at diagnosis, 2) therapies received, 3) assessments used in diagnosis, 4) Gross Motor Functional Classification System (GMFCS) level at age 2, 5) CP subtype, and 6) selected demographic factors. Charts were specifically reviewed for inclusion of three specific evidence-based assessments: GMA, HINE, and MRI.
Preliminary analyses were performed with data from 62 patients seen between 2010-2021. Data were divided into two cohorts (2010-2017, 2018-2021). MRI, GMA, and HINE were integrated more frequently in the later cohort (MRI: 27 to 48%, GMA: 0% to 79%, HINE: 0% to 17% of children). Mean age (corrected for prematurity) of CP diagnosis was comparable among groups (2010-2017: 15.4 ± 6.9 months, 2018-2021: 14.9 ± 8.2 months, independent samples t-test p =.782). Across all years, there was a difference in corrected age at diagnosis among children who received none (16.9 ± 7.7 months), one (16.2 ± 8.2 months), or two or more (10.2 ± 2.7 months) of the three evidence-based assessments (one-way ANOVA, p = .015). Analysis is ongoing with data through 2022 and will be completed prior to presentation. Additional planned statistical analyses will incorporate potential confounding variables (GMFCS level, demographic factors, social determinants of health) and other objective motor tests used. Analyses will also investigate the potential relationship between assessments used and timeline of access to rehabilitation therapies. As clinical settings increasingly implement these guidelines and the related assessment tools, this study will determine the subsequent real-world impact on diagnosis timelines and access to intervention for infants at risk for CP and may identify remaining gaps related to guideline implementation and feasibility.
Other
Optimizing Music-Based Interventions for Stroke Rehabilitation
1New York University, New York City, USA. 2University of Vienna, Vienna, Austria
Motor Rehabilitation
Development of a Biomechanical-based Classification System for Informing Precision Treatment of Post-Stroke Walking Impairment
1Ralph H. Johnson VA Heath Care System, Charleston, USA. 2Medical University of South Carolina, Charleston, USA. 3University of Texas, Austin, USA. 4Brooks Rehabilitation, Jacksonville, USA
Recovery of walking ability after a stroke continues to be highly variable because rehabilitation interventions do not always adequately address the complex heterogeneity of stroke pathology. Rehabilitation for post-stroke walking recovery has historically focused broadly on improving function, such as walking speed or distance walked. Unfortunately, measures of function can be influenced by a variety of impairments, like biomechanical or muscle coordination deficits. The lack of impairment information from functional measures naturally limits how well these measures can inform clinical decisions for designing and prescribing targeted precision rehabilitation therapies. The NIH StrokeNet Round Table has championed that there is a critical need to develop a measurement or classification system that is indicative of impairments post-stroke during walking to elucidate individual differences and advance precision treatment. Biomechanical deficits are an ideal candidate for informing a classification system because they are commonly present post-stroke and many individuals have multiple co-occurring deficits. Additionally, biomechanical deficits have been considered potential rehabilitation “biomarkers” because clinicians can deliver targeted interventions that are personalized to an individual’s presentation. The purpose of this study is to test whether biomechanical measures during post-stroke walking can be used to develop a classification system for individuals post-stroke. Our primary hypothesis is that biomechanical measures can differentiate individuals post-stroke into phenotypic groups (i.e., commonly co-occurring deficit clusters). Our secondary hypothesis is that performance across clinical measures will be normally distributed within subgroups and there will not be distinct group differences across subgroups.
Retrospective data is provided by a research database sponsored by the NIH Center of Biomedical Research Excellence in Stroke Recovery (IRB approval – Medical University of South Carolina). The dataset includes kinetic, kinematic, and clinical measures for 200 individuals with chronic stroke (> 6 months from diagnosis). Kinetic and kinematic data were collected during self-selected treadmill walking without an assistive or orthotic device. Various kinematic and kinetic measures were quantified using the time integral of each measure within each gait cycle region. Clinical measures include self-selected and fastest comfortable overground walking speed, 6-minute walk test, the lower extremity Fugl-Meyer, Berg Balance Scale, Functional Gait Assessment, Activities-specific Balance Confidence Scale, and Stroke Impact Scale. Cluster analysis statistical methods was used to determine the number and define the structure of subgroups. Descriptive statistics were used to define the clinical profile for each subgroup. Differences between clinical measures for each subgroup were tested with a one-way ANOVA and post-hoc t-tests with a Tukey-Kramer correction applied for multiple comparisons.
Results from this project are ongoing. Data and findings will be in-hand for presentation at the annual meeting. Findings are expected to help provide clinicians and researchers with a mechanistic framework for differential diagnosis and treatment of post-stroke walking impairment.
Stroke
The use of a gamified upper extremity rehabilitation system for in-clinic and at-home therapy facilitation
1Texas Biomedical Device Center, University of Texas at Dallas, Richardson, USA. 2School of Behavioral and Brain Sciences, University of Texas at Dallas, Richardson, USA. 3Erik Jonsson School of Engineering and Computer Science, University of Texas at Dallas, Richardson, USA. 4Baylor Scott & White Institute for Rehabilitation, Dallas, USA
The introduction of gaming technologies to physical rehabilitation has gained traction in recent years. They are applicable in the clinic and at home. Apart from increasing the amount of practice, they provide a seamless way to individualize therapy and monitor progress. The RePlay upper extremity rehabilitation system is one such technology that is easy to use and has diverse games and exercises. Here, we describe how we used this novel system to facilitate rehabilitation for individuals with upper extremity impairments due to stroke and spinal cord injuries.
The RePlay system consists of specialized controllers, games, and exercises that are designed to target specific impairments. We used the clinical assessment scores of the participants to determine the baseline exercise prescription and difficulty. That is, the Graded Redefined Assessment of Strength, Sensibility, and Prehension (GRASSP) for the spinal cord injury participants and the Upper Extremity Fugl-Meyer Assessment (UEFM) for the stroke participants. In addition, we used game performance metrics to determine exercise progression. Participants ranged in age from 18 to 64 years old, with chronic motor deficits caused by traumatic spinal cord injury that occurred at least 12 months before enrollment, and from 18 to 80 years old, with a chronic ischemic stroke that occurred at least 12 months before enrollment. They completed three 90-minute in-clinic therapy sessions per week for twelve weeks. After the in-clinic phase of the study, some participants continued therapy at home, where we tracked gameplay remotely and adjusted difficulty accordingly.
Fifteen spinal cord injury participants and seven stroke participants have completed therapy in the clinic. Of these twenty-three participants, eight are currently using the system at home. They completed hours of therapy with diverse exercises, and the difficulty progressed over time. The prescriptions for stroke and spinal cord injury were more different than the prescriptions for people with the same injury type. Also, the prescriptions for those with the same injury type and similar clinical assessment scores were more alike.
Overall, the study suggests that the RePlay system can be an effective way to facilitate physical rehabilitation for individuals with upper extremity impairments. It made it easy to provide individualized therapy to stroke and spinal cord injury patients. Ongoing studies are combining this approach with vagus nerve stimulation. The use of game-based systems as adjuncts to conventional therapy continues to show potential.
Motor Rehabilitation
Linking post-stroke neurophysiology to neuroanatomy: Novel method to extend voxel-lesion mapping to multi-dimensional EEG data
1Center for Neurotechnology and Neurorecovery, Department of Neurology, Massachusetts General Hospital, Boston, USA. 2VA RR&D Center for Neurorestoration and Neurotechnology, Department of Veterans Affairs Medical Center, Providence, USA. 3Carney Institute for Brain Science and School of Engineering, Brown University, Providence, USA. 4Harvard Medical School, Boston, USA. 5Department of Neurology, University of California, Los Angeles, USA. 6California Rehabilitation Institute, Los Angeles, USA. 7Institute of Medical Psychology and Behavioral Neurobiology, University of Tübingen, Tübingen, Germany. 8TECNALIA, Basque Research and Technology Alliance (BRTA), Neurotechnology Laboratory, San Sebastián, Spain
EEG measures of neural activity contain rich spatio-temporal information across different channels and frequency bands. EEG recordings after stroke have shown group-level changes in both resting state and task-evoked activity. However, the neuroanatomical basis of these EEG changes remains poorly understood with relatively few studies attempting to link the neuroanatomy of lesions to the changes seen in EEG signals. Here we present a novel method to analyze the relationship between patterns of neuroanatomical injury and changes in neural activity across different cortical regions and frequency bands.
EEG-Voxel Lesion Mapping (EEG-VLM) builds on voxel-lesion symptom mapping (VLSM). In VLSM a single t-map is created by comparing values for a single “symptom” (i.e. one value per patient) at every voxel between those patients who do and don’t have a lesion at that voxel. To incorporate the multi-dimensional nature of EEG signals (n channels x m frequencies), a separate t-map is first created for each pair (channel, frequency). Using adjacency maps (based on neighboring channels and frequencies), we designed a novel, tractable, clustering algorithm which finds maximal significant clusters (containing x voxels, y EEG channels and z frequencies). A cluster captures a relationship where lesions in any of the x voxels co-occur with significant changes at each EEG pair (channel, frequency). The significance of these clusters can then be assessed using permutation statistics.
We applied this method to a previously recorded dataset in which chronic stroke patients (N=30) with arm and hand motor impairments performed cued hand opening/attempted opening using their healthy or paretic hand. Preliminary findings show that during attempted paretic movement, EEG-VLM finds a significant cluster (permutation test, p<0.05) involving post-cue beta desynchronization over the lesioned hemisphere and the presence of lesions in frontal white matter.
EEG-VLM is a novel and unbiased method for relating neurophysiologic changes after stroke with neuroanatomic lesions. We propose that this method will enable an improved mechanistic understanding of stroke-induced changes in EEG signals and improve detection of behaviorally relevant EEG-biomarkers.
Stroke
Improving distal arm motor function in a chronic stroke survivor with intensive chopstick operation skill training in conjunction with tPBM: A Case Report
SUNY Upstate Medical University, Syracuse, USA
Transcranial photobiomodulation (tPBM) with near-infrared light is a potential non-invasive brain stimulation technique for stroke recovery. A case study was carried out to investigate the feasibility and long-term advantages of tPBM in conjunction with an arm and hand motor skill training program for the improvement of distal arm motor function in chronic stroke survivors. A 73-year-old individual with chronic stroke and mild upper extremity motor impairment (FMA-UE = 54/66) participated in a 4-week motor training program coupled with tPBM. The individual completed eight training sessions, each consisting of a 1-hour intensive practice of chopstick operating skills followed by a 15-minute tPBM treatment. The participant utilized a set of chopsticks to pick up and move various objects from one plate to another at a time during the chopstick operation skill practice. During early sessions, the participant used a chopstick helper, which grips the proximal ends of the chopsticks and makes them easier to use. The tPBM was delivered to the scalp over the ipsilesional sensorimotor areas at a dosage of 20 J/cm2 using a 22.48 cm2 probe that produces near-infrared light. The distal arm and hand function was measured using the time scores of the Wolf Motor Function Test distal item subset (dWMFT), Box and Block Test (BBT), and modified Box and Block Test (mBBT) at baseline, every week before the training session, and at the conclusion of the final training session. The Motor Activity Log (MAL) was also used to assess the usage of participants’ paretic arms in daily activities. Finally, chopstick operation skill was assessed using kinematic measures of chopstick performance. The participant completed the four-week training program without incident. The participant improved in dWMFT (baseline average = 5.54 seconds; post-training average = 3.82 seconds), MAL Quality of Use (baseline average = 3.63; post-training average = 4.35), BBT (baseline paretic hand = 39; post-training paretic hand = 45), and mBBT (baseline paretic hand = 17; post-training paretic hand = 21) after completing the 4-week training session. At week three, the participant was able to operate the chopstick skill without the chopstick helper. However, he did not exhibit substantial improvement in the kinematics of chopstick performance. The participant stated during the program’s exit interview that he feels more confident in utilizing his paretic arm and hand in daily activities. He also stated that, he is more engaged in-home exercises as a result of training. This case study demonstrates that a 4-week arm and hand motor skill training program combined with tPBM is feasible and can improve distal arm motor function in chronic stroke survivors. More large-scale clinical trials will be needed to assess the long-term benefits of tPBM in conjunction with a motor skill training program.
Stroke
The Use of Automatic Closed-loop Vagus Nerve Stimulation During Rehabilitation For Stroke or Spinal Cord Injury
1Texas Biomedical Device Center, Richardson, USA. 2Erik Jonsson School of Engineering and Computer Science, Richardson, USA. 3School of Behavioral and Brain Sciences, University of Texas at Dallas, Richardson, USA. 4Baylor Scott and White Institute for Rehabilitation, Dallas, USA
Many stroke and spinal cord injury survivors are left with chronic motor disability. Paired VNS was recently approved by the U.S. Food and Drug Administration for the treatment of upper extremity motor deficits associated with chronic ischemic stroke. Three clinical trials indicate that vagus nerve stimulation (VNS) combined with supervised physical rehabilitation after stroke leads to increased recovery of upper limb motor function, but additional performance gains are not observed with continued unsupervised therapy. Development of tools to increase repetitions and enhance VNS timing may lead to improved outcomes. To address this, we developed RePlay, a novel mobile system to facilitate unsupervised rehabilitation exercises, and incorporated a dynamic algorithm that can automatically trigger stimulation based on selected parameters of movement.
RePlay leverages accessible technology to provide a simple tool that allows users to perform common rehabilitative exercises in a gameplay environment. RePlay collects quantitative time-series force and movement data from handheld devices and inputs the information to the dynamic algorithm for real-time VNS triggering based on selected parameters of movement. The dynamic algorithm produces reliable triggering near a desired interval while triggering during movements of the largest amplitude. The dynamic algorithm continually adapts over time to adjust for intermittent periods of rest and person-to-person variability and can flexibly be applied to signals from various controllers (handheld sensors, touchscreen, and keyboard) while maintaining high triggering selectivity.
The prospective benefits of RePlay are being studied in two ongoing double-blind randomized placebo controlled clinical studies in chronic stroke and spinal cord injury participants. During the long-term follow up, participants are continuing in open-label follow-on studies with at-home therapy. Preliminary evidence indicates RePlay combined with closed-loop VNS may reduce upper limb deficits in spinal cord injury. The RePlay system and its dynamic algorithm represent novel tools to facilitate access to unsupervised rehabilitative exercises with closed-loop neuromodulation that may improve overall patient outcomes.
Motor Rehabilitation
Investigating the relationship between anatomical and physiologic measures of the corticospinal tract and upper extremity motor function after acute stroke
1MGH Institute of Health Professions, Boston, USA. 2MGH Center for Neurotechnology and Neurorecovery, Massachusetts General Hospital, Boston, USA. 3Massachusetts General Hospital, Department of Neurology, Boston, USA. 4VA RR&D Center for Neurorestoration and Neurotechnology, Providence VA Medical Center, Providence, USA
The initial weeks after stroke are critical for upper extremity (UE) motor recovery. Both anatomical and physiologic changes occur in the brain following a stroke and influence the degree of recovery achieved. While it is known that integrity of the corticospinal tract (CST) plays an important role in UE motor function and recovery post stroke, the relationships between CST anatomy, physiology, and UE motor function early after stroke are not well understood. Therefore, this study aims to investigate the association between assessments of CST anatomy via magnetic resonance imaging (MRI), CST physiology via transcranial magnetic stimulation (TMS), and UE motor function during acute stroke recovery. We hypothesized that due to network effects after acute stroke (i.e., diaschisis), physiologic measures of CST would be more strongly related to UE motor function than anatomical measures when assessed during acute stroke hospitalization. Thirteen adults (65.44 ± 12.52 years) with first-ever stroke resulting in UE weakness were recruited and enrolled from the inpatient neurology service at Massachusetts General Hospital as part of the Stroke Motor reHabilitation and Recovery sTudy (SMaHRT), an ongoing, prospective, observational cohort study. TMS measures and UE motor assessments, including the Fugl-Meyer Assessment-Upper Extremity (FMA-UE), Box and Block Assessment (BB), Grip strength, and SAFE Score were performed at the hospital bedside. The resting motor threshold (RMT) of the lesioned hemisphere was used as a measure of CST physiology. CST injury was used a measure of CST anatomy and derived from MRIs obtained as part of the standard-of-care acute stroke workup. Stroke lesions were manually traced, binarized, finalized and spatially transformed into Montreal Neurological Institute and Hospital (MNI) stereotaxic space. CST injury was estimated using maximum area overlap between the stroke lesion and binarized CST tract. TMS was performed within 5.2 ± 2.2 days post-stroke. RMT and CST injury demonstrated a weak positive relationship (rho=0.37, p>0.05). RMT showed strong, significant negative associations with all UE motor behavioral outcomes (FMA-UE: rho=-0.71, p<0.01; BB: rho=-0.78, p<0.01; Grip Strength: rho=-0.72, p<0.01; SAFE Score: rho=-0.78, p<0.01) while CST injury did not demonstrate significant associations with any of the UE behavioral measures (FMA-UE: rho=-0.35; BB: rho=-0.32; Grip Strength: rho=-0.26; SAFE Score: rho=-0.26; all p>0.05). These results indicate that TMS- and imaging- based assessments have different relationships to UE motor function during the acute stroke recovery period. TMS neurophysiologic measures are more strongly associated with initial UE deficits than CST anatomical measures, suggesting that TMS measures may capture the influence of network-level processes that are distinct to early stroke recovery, such as diaschisis, on UE motor function. Future directions will focus on longitudinally assessing the relationship between CST anatomy, physiology, and UE motor behavior to better develop UE recovery prediction models that an inform personalized stroke rehabilitation.
Motor Rehabilitation
Reduced cortical sensory processing during whole-body motion perception after stroke
1Emory University School of Medicine, Atlanta, USA. 2Emory University, Atlanta, USA. 3Georgia Institute of Technology, Atlanta, USA
Somatosensory perceptual deficits are common after stroke but their role in balance function is unclear. Balance control becomes less automatic and more cognitively demanding in individuals with lower balance ability. We hypothesize that this shift influences how the brain integrates sensory information for perception of whole-body motion (WBM). Cortical activity involved in balance control can be quantified using electroencephalography (EEG). The balance N1 is a characteristic EEG event related potential over sensorimotor cortical areas that indexes a sensory driven error signal. The N1 attenuates with experience and prior sensory stimulation, and scales with balance ability, but its functional role in WBM perception is unknown. We predict that individuals with stroke will have reduced N1 responses during WBM perception compared to controls, and will be associated with lower WBM perception and balance ability.
WBM perception was tested in 16 individuals with chronic stroke and 24 healthy adults (18 younger adults (YA), 6 older adults (OA)) while recording EEG continuously. We used a 2-alternative forced choice paradigm in which participants reported whether pairs of support-surface perturbations were in the “same” or “different” direction. The first perturbation was delivered in the backward direction, while the second perturbation was in the backward direction but deviated laterally at an angle that preferentially loaded the paretic or nonparetic leg. Responses were fit to a psychometric curve to quantify perceptual ability. Balance ability was quantified via the distance traversed on a narrowing beam. The N1 was defined as the first most negative potential 100-300 msec post-perturbation. We quantified N1 amplitude for each perturbation, and N1 modulation within the pair.
Individuals with stroke had reduced N1 amplitude compared to OA and YA (OA: p= 0.05, YA: p< 0.001). While N1 amplitude slightly reduced within a perturbation pair in stroke, the magnitude of attenuation was less than that of controls (OA: p= 0.005; YA: p< 0.001). Individuals with stroke had lower WBM perceptual ability compared to controls (YA: p< 0.001), which was preferentially lower when loading the paretic side compared to the nonparetic side (p=0.006) and associated with balance ability (paretic: r= -0.62, p= 0.01). There were no significant associations between N1 amplitude or modulation and WBM perceptual or balance ability.
Reduced N1 amplitude and reduced N1 modulation in stroke suggests deficits in processing sensory information during standing balance perturbations. Reduced N1 modulation is consistent with known sensory gating deficits in stroke. However, the lack of associations with perceptual ability highlights a distinction between initial sensory processing versus higher order top-down cognitive control processes involved in perception. Since WBM perception was impaired in stroke and associated with balance dysfunction, investigating other cortical signatures of WBM perception will be important for novel and personalized rehabilitation interventions.
Stroke
Learning New Gait Patterns after Stroke: Do Stroke Survivors with Mild Motor Impairments Exhibit Deficits in Learning?
1Robotics Department, University of Michigan, Ann Arbor, USA. 2Physical Medicine and Rehabilitation, Michigan Medicine, Ann Arbor, USA. 3Department of Biomedical Engineering, Wayne State University, Detroit, USA. 4Department of Chemistry, University of Michigan, Ann Arbor, USA. 5Department of Kinesiology, Michigan State University, Lansing, USA. 6Department of Mechanical Engineering, Michigan State University, Lansing, USA. 7Department of Biomedical Engineering, University of Michigan, Ann Arbor, USA. 8Department of Kinesiology, University of Michigan, Ann Arbor, USA
Stroke is a leading cause of adult disability, with many stroke survivors experiencing difficulty in walking even years after stroke. It is believed that motor learning mechanisms are active during the recovery period and interact with stroke rehabilitation. Thus, it is critical to understand the extent of motor learning deficits to develop effective rehabilitation paradigms. Previous studies on upper extremities have shown that stroke typically affects the processes underlying motor control and execution but not the learning of those skills. However, it is unclear if this extends to a more functional lower-extremity task and whether individuals with minimal motor impairment also exhibit such deficits. Therefore, this study aimed to investigate leg motor skill learning during walking in stroke survivors and compare the results to neurological intact older adults.
Thirty participants (10 stroke; 20 controls) were tested on a treadmill over two consecutive days using a foot-trajectory tracking task. On day 1, participants learned a new gait pattern by performing a task that necessitated greater (1.3×) hip and knee flexion during the swing phase of the gait. On day 2, participants repeated the task with their training leg to test retention. Two test blocks (pre-test and post-test) and eight training blocks were performed on both days. Each block consisted of one minute of practice and one minute of rest. An average tracking error was computed for each block to determine task performance. Tracking error was calculated as the difference in the area (in pixels) between the participant’s actual trajectory and the target trajectory for each stride. This stride-by-stride tracking error was then normalized to the target template and averaged across strides for each block. An analysis of covariance with block (pre-test and post-test) as the within-subjects factor, group as the between-subjects factor, and pre-test block as a covariate was used to evaluate differences between groups.
The results showed that although stroke survivors were able to improve their tracking performance, the amount of learning in stroke survivors was lower in comparison with the control group on both days (Day 1: Δ = 3.1±1.5% vs. 6.9±1.0%; p=0.05; Day 2: Δ = 1.6±1.1% vs. 4.3±0.8%; p=0.046). This suggests that even high-functioning stroke survivors may have difficulty acquiring new motor skills related to walking, which may be related to the underlying neural damage caused at the time of stroke. Furthermore, it is likely that stroke survivors may require a longer time to acquire new motor skills related to walking.
In conclusion, this study highlights the need for further research on motor skill learning in stroke survivors and suggests that rehabilitation programs for stroke survivors should be designed to specifically target motor skill learning to improve outcomes after stroke.
Motor Rehabilitation
The impact of the COVID-19 pandemic on rehabilitation delivery and outcomes in the province of Quebec
1McGill University, Montreal, Canada. 2Centre for Interdisciplinary Research in Rehabilitation of Greater Montreal (CRIR), Montreal, Canada. 3Jewish Rehabilitation Hospital, CISSS-Laval, Laval, Canada. 4Villa Medica Rehabilitation Hospital, Montreal, Canada. 5University of Montreal, Montreal, Canada. 6Institut de réadaptation Gingras-Lindsay-de-Montréal, CIUSSS-CSMTL, Montreal, Canada. 7University of Quebec at Chicoutimi, Saguenay, Canada. 8Laval University, Quebec, Canada. 9Center for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRIS), Quebec, Canada. 10University of Sherbrooke, Sherbrooke, Canada. 11Centre de recherche sur le vieillissement, CIUSSS de l’Estrie-CHUS, Sherbrooke, Canada
Standardized post-stroke rehabilitation protocols aim to optimize motor and cognitive functions(1). However, the COVID-19 pandemic disrupted the delivery of rehabilitation services in Quebec, where centers underwent a reorganization in delivering services, especially during the first two waves of the pandemic(2).
This study aims to estimate the impact of the pandemic on the delivery of rehabilitation care to stroke users having contracted or not COVID-19 by (1) comparing their healthcare indicators and rehabilitation outcomes and (2) identifying factors of influence. We hypothesized that rehabilitation outcomes would be negatively affected in all stroke users during the pandemic due to operational changes in the provision of rehabilitation services, even more so for those who contracted COVID-19 because of their concomitant health condition.
Three post-stroke groups were included: (i) with (COV+: 96), (ii) without (COV-: 102) admitted to COVID-19-designated rehabilitation centers in Quebec during the first two waves (March 2020/21), and (iii) pre-COVID users (preCOV: 98) admitted the year before. Demographics and descriptive information regarding acute and rehabilitation care (length of stay (LOS), intubation, physical/occupational therapy (PT/OT) sessions) were extracted from medical records. The primary rehabilitation outcome was the Functional Independence Measure (FIM)(3). The average increase in the FIM score per day (Rehabilitation Efficiency-REy)(4) was calculated. Kruskal-Wallis and Chi-Square tests compared these variables between groups. Factors influencing the FIM scores at discharge were also identified using multiple linear regression models.
COV+ users were older than COV- and preCOV (76±11/73±11/74±10 years, p<0.03). They had longer stay in acute care (30±21/19±12/20±16 days, p<0.001), with a higher proportion being intubated (19%/9%/3%, p<0.03). COV+ users also had longer rehabilitation stay (65±38/45±30/52±31 days, p>0.02), more frequent relapses (33%/13%/21%, p<0.003), lower FIM scores on admission (FIMpre: 68±23/76±22/75±25, p<0.04) and at discharge (FIMpost: 92±29/104±20/101±21, p<0.02), and received more PT/OT sessions (PT: 35±23/25±17/31±24, p<0.03, OT: 33±22/24±16/29±22, p<0.03). COV- users had higher REy scores (0.8±0.6) than preCOV ones (0.6±0.4, p<0.024) and a shorter LOS (45±30/52±31 days, p>0.04). A higher REy score and FIMpre score positively influenced rehabilitation outcomes, whereas a longer stay in acute care was detrimental to recovery.
As hypothesized, post-stroke COV+ users showed diminished healthcare status at rehabilitation admission and made less functional progress despite having longer stay and more PT/OT sessions. REy was unexpectedly high among COV- users, with a shorter stay in rehabilitation compared to the preCOV users, despite similar profiles to pre-COV group. Post-stroke COV+ population may need more tailored in-patient rehabilitation care. Factors underlying these differences will require further exploration, but these results already provide insight for delivering rehabilitation during a pandemic.
Stroke
Contralateral fMRI activation for line bisection judgments after right-hemisphere stroke
Georgetown University Medical Center, Washington, DC, USA
Cognitive/Language Rehabilitation
Effects of Hyaluronidase Injections on Neural and Non-Neural Muscle Stiffness Post Stroke
1Johns Hopkins University, Baltimore, USA. 2Northwestern University, Illinois, USA
Stroke
Test-retest reliability and measurement error of spatial-temporal measures of movement variability in finger coordination task
1Department of Kinesiology, Doctor of Physical Therapy Program, Storrs, USA. 2Department of Physical Therapy, University of North Georgia, Dahlonega, USA
Motor Rehabilitation
A Preliminary Study of Motor Control Abnormalities in the First 3 Months After Stroke
1University of Rochester, Rochester, USA. 2Department of Neurology, University of Rochester, Rochester, USA
Stroke
Input-output property of soleus short latency crossed spinal inhibition in people with chronic incomplete spinal cord injury
Medical University of South Carolina, Charleston, USA
Stimulating the tibial nerve in the ipsilateral soleus produces temporary suppression of ongoing EMG activity in the contralateral soleus with latencies of spinal origin[1]. This short-latency crossed spinal inhibition (CSI), presumably mediated by muscle afferents, could be a useful tool for examining spinal mechanisms of interlimb coordination[2]. The earlier study[1] found, in adults (31±10 yrs) with no neurological conditions, larger CSI with larger M-wave in ipsilateral soleus (i.e., higher tibial nerve stimulus intensities). Currently it is unknown whether and to what extent such input-output relation exists in people after spinal cord injury (SCI). Thus, in hope to better understand spinal mechanisms of spastic motor impairments after incomplete SCI, this study examined soleus CSI at various stimulation intensities bilaterally in people with and without chronic incomplete SCI.
Ten adults (57±16 yrs in age) with spasticity due to chronic (2-50 yrs post) incomplete SCI and 10 age-matched individuals (55±14 yrs) with no known neurological conditions participated in this study. EMG signal was recorded from the soleus and tibialis anterior (TA) bilaterally. To elicit the H-reflex and M-wave in the ipsilateral soleus and measure CSI in the contralateral soleus, the ipsilateral tibial nerve was stimulated in the popliteal fossa with a 1-ms square pulse when the standing participant had maintained soleus EMG activity within a pre-determined range (i.e., one’s natural standing level) for at least 2 s. Stimulus intensity was varied from soleus H-reflex threshold to the maximum H-reflex (Hmax) to an intensity just above what was needed to elicit the maximum M-wave (Mmax). CSI, typically observed around +6 ms from the H-reflex onset, was averaged in 4 separate stimulus intensity levels (i.e., <25%, 25-50%, 50-75%, and 75-100% ipsilateral soleus Mmax) and expressed as %background EMG (bEMG). A linear mixed model analysis was used to access the effects of stimulus intensity on the amount of inhibition in the more spastic leg and the less spastic leg of participants with SCI, and in the larger Hmax (expressed in %Mmax) leg and the smaller Hmax leg of non-SCI.
In non-SCI, stronger stimulation produced more robust CSI, in general. However, the effect of stimulus intensity was statistically significant in the larger Hmax leg (p=0.03) but not in the smaller Hmax leg (p=0.12). In individuals with SCI, the effect of stimulus intensity was significant in the less spastic leg (i.e., more CSI with stronger stimulation, p=0.001) but not in the more spastic leg (p=0.73). The lack of stimulus intensity-dependent modulation of CSI in the more spastic leg of individuals with incomplete SCI suggests altered CSI input-output property after SCI. Altered CSI may be a mechanism of multifaceted abnormal spinal inhibition and resulting sensorimotor function impairments in this population.
Spinal Cord Injury (SCI)
Standing posture improves upper-limb sensorimotor performance on a robotics-based task with high proprioceptive feedback demands
1Emory University, Atlanta, USA. 2Georgia Institute of Technology, Atlanta, USA
Upper-limb therapy/research is largely conducted while seated, however, a substantial proportion of real-world activities of daily living take place while standing. There is a dearth of research into the impact of posture on upper-limb performance. It is unclear how posture impacts performance in a continuous task with greater demands on sensory feedback. Here, we present preliminary results from neurotypical younger adults as part of an ongoing study into the effects of proprioceptive deficit due to aging and stroke on sensorimotor behavior and neural systems.
The effect of posture on performance was tested in 12 neurotypical young adults (ages 18-26). We utilized a custom target tracking task (TT) built for the Kinarm end-point robot (Ontario, Canada). The TT requires participants to move a manipulandum to match a moving target’s position (2cm circle) along a circular trajectory. Trial performance was the average distance between the manipulandum and the center of the target during 10 consecutive seconds of tracking (144 total trials). In addition to posture (seated vs. standing), we tested the effect of sensory feedback (vision vs. no vision of the limb) on performance, with the no vision condition requiring reliance on proprioception to track the limb.
Participants demonstrated significantly greater error when performing without visual feedback of the limb. Error was significantly reduced when standing, though only in the no-visual feedback condition (p < 0.001). Similar significant effects were observed for both the dominant right limb (p < 0.001) and non-dominant left limb (p < 0.001).
Neurotypical younger adults performance improved while standing during a sensorimotor task that required continuous proprioceptive feedback. Whereas prior research utilizing a discrete reaching task reported worse performance when standing. This suggests that posture impacts upper-limb performance differently based on task features (discrete vs. continuous) and available sensory information (vision vs. proprioception). Our findings may have implications for upper-limb research and therapy, for instance, clinicians may choose to alter posture during therapy based on task demands and sensory function of the patient.
Other
Using sensory stimulation to enhance neuroplasticity in the sensorimotor cortex in stroke survivors to promote upper limb motor recovery
1Medical University of South Carolina, Charleston, USA. 2Ralph H. Johnson VA Healthcare System, Charleston, USA
Stroke
Estimating transfer of motor skill learning post-stroke from a large sample “in the wild” practice data
1University of Southern California, Computer Science, Los Angeles, USA. 2University of Southern California, Biokinesiology and Physical Therapy, Los Angeles, USA. 3UC Irvine, Department of Mechanical and Aerospace Engineering, Irvine, USA. 4Flint Rehab, Irvine, USA
Choosing specific motor tasks to practice during motor therapy following a stroke affecting the upper extremity is typically determined by clinicians based on their experience. However, some tasks may be more beneficial than others, as they could lead to improvements that generalize across other tasks or body parts. Here, we study the possible transfer of learning among K=8 tasks for N=366 individuals who followed the FitMi training at home for up to 80 weeks. We selected eight tasks that span the different arm and hand joints, including two hand, two wrist, one elbow, one shoulder, and two reaching tasks (lateral and forward). The individuals form a subset of FitMi users who have practiced all eight tasks for at least two weeks.
We modified a previous hierarchical Bayesian dynamical model of outcomes post-stroke as a function of dose (Schweighofer et al. 2022). Task performance was taken as the average number of repetitions per second each week. Learning dynamics were modeled using a first-order state-space model with a K-dimensional state encoding a patient’s performance for each task. The dose input was taken as the number of repetitions (for a task) each week. The model was parameterized by K decay rates, in the form of a diagonal transition matrix, and K2 learning rates, in the form of a full K×K input matrix. To avoid spurious discoveries, we imposed a horseshoe prior to enforce a sparse learning rate solution. We report the hierarchical (i.e., population) learning rate matrix to compare the effects of practicing each task (diagonal terms) and the effect of practicing a task on the performance of another task (off-diagonal terms). Model fit was examined using RMSE (on both population and subject levels), and model comparison was performed via Bayesian cross-validation. Parameter estimation was conducted for all participants simultaneously via Stochastic Variational Inference (SVI) in NumPyro.
Motor Rehabilitation
Cortical and functional changes in Hand Function after 3-weeks of Training Using a Novel Passive Device
1Rotman Research Institute, Baycrest Hospital, Toronto, Canada. 2IRegained Inc, Sudbury, Canada. 3Department of Kinesiology, McMaster University, Hamilton, Canada. 4School of Kinesiology, Lakehead University, Thunder Bay, Canada. 5School of Rehabilitation Science, McMaster University, Hamilton, Canada
Stroke is the leading cause of adult disability with ~ 60% of persons with stroke (PWS) experiencing long-term hand function impairments. As hand function plays an important role in performing activities of daily living, PWS face a significant impact in their quality of life. This study explored the impact of a novel passive hand function therapy (HFT) device - the MyHandTM System, to address finger and hand disability. The device allows for repetitive individual finger activity and/or various grasp types to be performed in a gamified environment. The primary objective of this study was to explore the efficacy of the MyHandTM System, using a 3-week training program designed to improve sensorimotor function through a granular, progressive approach involving gamified video feedback. Changes in cortical activity and hand function pre and post intervention was estimated using transcranial magnetic stimulation (TMS) and functional assessments. Eleven chronic (123.27 [CI 95%: 45.19 to 201.36] months post-stroke) PWS participants (59.81 [CI 95%: 46.17 to 73.47] years of age) participated in this study. The intervention consisted of fifteen, 1-hour HFT sessions over three-weeks using the MyHandTM System, for the duration of the study, with participants taking breaks as needed during sessions. The primary outcome measure was the Action Research Arm Test (ARAT), with the Box and Block Test (BBT) and the ABILHAND questionnaire as secondary outcome measures. All assessments were performed pre and post intervention. The post intervention was done within 5 days of the final intervention session. Changes in cortical excitability (estimated with motor evoked potential at the first dorsal interosseous) and inter-hemispheric inhibition were recorded using TMS pre and post intervention. No assessments, whether functional or TMS were completed the same day as treatment. Significant functional improvements were observed in the mean scores for ARAT (PRE: 19.36 [CI 95%: 17.41 to 21.31], POST: 26.27 [CI 95%: 24.32 to 28.22], p = .004) and BBT (PRE: 6.55 [CI 95%: 4.77 to 8.32], POST: 10.18 [CI 95%: 8.41 to 11.96], p = .023). These changes met the minimal clinically important difference (MCID) criteria. ABILHAND logit score improved to fulfill the MCID criteria (PRE: 0.58 [CI 95%: 0.27 to 0.89], POST: 0.95 [CI 95%: 0.64 to 1.26]), however, demonstrated marginal statistical significance (p = 0.078). Data pertaining to cortical changes are presented only for one patient representing the neurophysiological changes as estimated by TMS. Results from this study indicate that fully passive HFT devices can increase function and reduce impairment of the hand in chronic PWS. We posit that the functional outcome measure data along with TMS data support this efficacy of the MyHand System. Studies that further investigate passive options for HFT in PWS should continue with larger sample sizes and employ practical timelines for neurorehabilitation.
Stroke
You don’t have to be at risk of falling to be afraid of falling: Examining the relationship between fear of falling and balance impairment at inpatient discharge in ambulatory stroke survivors
MGH Institute of Health Professions, Boston, USA
Stroke
The value of dynamic grip force modulation as a potential biomarker for hand function recovery following stroke
Department of Health Sciences and Research, College of Health Professions, Medical University of South Carolina, Charleston, USA
Precise modulation of grip force is vital in daily life hand function. However, while up to 80% of stroke survivors experience impairment of upper extremity sensorimotor function, standard clinical diagnostic procedures in stroke rehabilitation largely ignore dynamic grip force modulation. Furthermore, we still need a better understanding of the relationship between individual components of dynamic force modulation (e.g., force increase or release) and daily life functions. To fill this knowledge gap, we tested uni- and bimanual dynamic grip force adjustment in volunteers recovering from stroke (inclusion criterion: at least voluntary release of finger flexor activity) and neurologically healthy volunteers using a custom-designed isometric visuomotor force tracking task. We extracted separate outcome parameters from the grip force profiles describing distinct components such as a) speed of force increase and b) decrease, as well as c) median precision and d) variability of constant force production.
We used a supervised machine learning algorithm (support vector machine, SVM, with k-fold cross-validation) for binary classification of groups (stroke versus control group), task conditions (uni- versus bimanual), and to quantify the active range of motion evaluated with upper extremity Fugl-Meyer Assessment (UEFMA) within the stroke group alone.
Based on our preliminary data (stroke group N=22, control group N=18), we found very high classification accuracy (90%) in correctly identifying patients when including duration of force increase and force as predictor variables. Adding the median precision and variance of constant force production increased the accuracy of correctly identifying control group members (55% accuracy). However, it did not change the accuracy of classifying members of the patient group.
While bimanual force modulation increased the overall rate of failed attempts in the patient group, dynamic bimanual force modulation showed higher variability even in higher-functioning patients (UEFMA >50). Furthermore, duration of force increase (slower force increase with higher UEFMA score) and median force precision (higher precision with higher UEFMA score) of the affected hand in the uni- and bilateral conditions showed associations with UEFMA to varying degrees, with a stronger association in the bilateral condition.
Together these preliminary findings of our ongoing study suggest that dynamic components of whole-hand grip force modulation may serve as a sensitive biomarker of recovery of hand function. Finally, the differentiation of dynamic force modulation of the paretic hand in the uni- in contrast to the bimanual condition can potentially add clinically relevant information about sensorimotor capacity.
Stroke
Interactions between spatial navigation ability and cognitive function in the aging brain
1Emory University, Atlanta, USA. 2Georgia Institute of Technology, Atlanta, USA
Cognitive/Language Rehabilitation
Restoration of Mobility and Balance in People with Secondary Progressive Multiple Sclerosis: A Case Series
MGH Institute of Health Professions, Boston, USA
Multiple Sclerosis (MS)
Forearm Postural Diversity and Complexity: Targets for Wearable Feedback after Stroke?
University of California, Irvine, Irvine, California, USA
Over 80 percent of people who have experienced a stroke incur upper extremity (UE) impairment resulting in reduced arm use in daily life. A few studies have examined the use of wearable feedback of the daily quantity of arm movement in order to promote recovery, but with limited success. We posit that it may be more effective to encourage an increase in beneficial patterns of movement, rather than simply the overall amount of movement each day. As a first step toward this goal, here we sought to identify statistical characteristics of daily arm movements that become more prominent as arm impairment decreases.
Using data obtained from a wrist IMU (i.e., the Manumeter) worn by 20 chronic stroke participants throughout their day, we identified several measures that increased as UE Fugl-Meyer (UEFM) score increased: forearm speed, forearm postural diversity (quantified by kurtosis of the tilt-angle), and forearm postural complexity (quantified by sample entropy of tilt angle). We grouped the participants based on their UEFM scores into severe, moderate, and mild impairment groups. We found that postural diversity and complexity most effectively distinguished groups (Cohen’s D = 1.1 and 0.99, respectively).
Based on these findings, we posit that encouraging people to achieve more forearm postural diversity and complexity might be therapeutically beneficial. But what exercises could one suggest to help patients achieve this goal? We recruited 8 unimpaired individuals and evaluated a set of 12 therapeutic activities for postural diversity and complexity. The activities were performed while seated and included: a set of conventional rehabilitation therapy exercises for the hand and arm, ping-pong, the card game (Speed and Klondike), American Sign Language, Tai-chi, Cornhole, Nintendo Switch sports (Chambara and Tennis), balloon volleyball, cup stacking, and FitMi exercises, where FitMi is a commercial sensor system that guides UE exercises. The participants performed each exercise for 10 minutes while wearing a wrist accelerometer, and we computed sample entropy and kurtosis of tilt angle with sliding windows of 1, 3, and 5 minutes. Engaging in conventional rehabilitation therapy exercises created high values for forearm postural diversity but not complexity. Playing the card game Speed and exercising with FitMi produced the highest values for both postural diversity and complexity.
These results suggest that quantitative measures of the diversity and complexity of forearm posture may be useful in wearable feedback systems for post-stroke rehabilitation. Patients could be encouraged to practice specific exercises to increase these measures.
Stroke
Control of interaction torques during single-joint arm movements in stroke survivors
1University of Southern California, Biokinesiology and Physical Therapy, Los angeles, USA. 2University of Southern California, Biomedical Engineering, Los angeles, USA. 3Tokyo institute of technology, Tokyo, Japan. 4Casa Colina Hospital and Centers for Healthcare, Pomona, USA. 5University of Southern California, Occupational Science and Occupational Therapy, Los angeles, USA
Sensory-motor strokes abruptly impair motor control of goal-directed movements. Generating fast, smooth and straight trajectories for multi-joint, planar reaches requires adequate compensation for the interaction torques that arise at both proximal and distal joints. It has been proposed that stroke survivors exhibit deficits in the control of these interaction torques (Beer et al., 2000). Here, we study to what extent stroke survivors activate shoulder muscles to control for anticipated interaction torques during fast movements restricted to the elbow. Neurotypical individuals learn to compensate for these torques by activating anticipatory shoulder muscles proportionally to the interaction torques (Gribble and Ostry, 1999). We hypothesize that, compared to the less-affected arm, the more-affected arm of stroke survivors will exhibit a diminished scaling between the interaction torques generated at the shoulder and the Electromyographic (EMG) activity of the Posterior deltoid (PD) and Pectoralis major (PEC) during fast single-joint elbow extension.
We recruited 26 (13 females), moderately to mildly impaired (UE-FM 42.8 ± 1.9), supra-tentorial chronic stroke survivors and tested their more- and less-affected arms. We studied the relationships between the interaction torques at the shoulder (computed from arm morphometry and shoulder and elbow angle, angular velocity, and angular acceleration) and the PD EMG activity. In addition, we cross-correlated the shoulder EMG activity S_EMG = PEC – PD, to the interaction torque signal at the shoulder. S_EMG preceded interaction torque signal more for the less affected side than for the more affected side. The PD EMG amplitude was less correlated with interaction torque on the more-affected side than the less-affected side.
Our results indicate that stroke survivors have a reduced ability to predictively compensate for interaction torques and to generate adequate anticipatory muscle activities. In future work, we will use structural imaging (DTI) to seek the neural correlate of this impairment. Because feedforward motor commands are thought to be learned via the cerebellum, damage in the cerebro-cerebellar pathways may preclude stroke survivors from updating their feedforward controllers.
Stroke
Short-latency spinal reciprocal inhibition in individuals with post-stroke hemiparesis
1University of Nevada, Las Vegas, Las Vegas, USA. 2Medical University of South Carolina, Charleston, USA
The short-latency reciprocal inhibitory pathway has been well examined in the non-impaired nervous system. Electrical stimulation of a mixed muscle nerve induces reciprocal inhibition (RI) of ongoing electromyography (EMG) activity in the antagonist muscle. The magnitude of inhibition increases with increase in intensity of stimulation. Individuals with post-stroke hemiparesis exhibit the inability to adequately activate paretic tibialis anterior (TA) muscle, which is often accompanied by inappropriate antagonist plantarflexor activity that would hinder toe clearance during the swing phase of gait. Thus, the problem of footdrop is attributable to reduced spinal RI between the ankle dorsiflexor and plantarflexor. However, this short-latency spinal RI has not been systematically examined in the stroke-impaired nervous system. The purpose of this study was to characterize the spinal RI in paretic and non-paretic legs of the individuals with chronic post-stroke hemiparesis.
Seven individuals with chronic post-stroke hemiparesis (53.8±15.4 years old; 10.1±6.1 years post-stroke) and 5 age-similar non-impaired individuals (42.7±11.1 years old) participated in the study. EMG was recorded from the TA and soleus (SOL) muscles in each leg. During standing, the common peroneal nerve (CPN) was electrically stimulated from below threshold to the maximal M-wave (Mmax) level to obtain a TA recruitment curve. Then, 25 pulses were delivered to the CPN at intensities that produced 25%, 50% and 75% of the TA Mmax. SOL EMG was rectified and RI was measured over a 7ms period (typically 40-50 ms post stimulus) including the peak suppression. The magnitude of RI was calculated as the difference in EMG amplitude between the 7ms inhibition period and the 30ms prestimulus period, and expressed as % prestimulus EMG. Negative values denote inhibition and positive values indicate facilitation. In both non-impaired legs, magnitude of RI increased linearly with increasing intensity of CPN stimulation (Left legs = -36.1%, -46.1%, -56.2%; Right legs = -35.4%, -47.5%, -51.8%, at 25%, 50%, 75% TA Mmax respectively). In both the paretic and non-paretic legs, on average, the magnitude of RI was smaller than that in the non-impaired legs. Additionally, the magnitude of RI did not consistently increase linearly with increasing intensity of CPN stimulation, and had greater variability (Paretic legs = -11.8%, -12.6%, -14.9%; Non-paretic legs = -25%, -31.6%, -36.1%). In one paretic leg, we observed facilitation in the SOL EMG under all intensities of CPN stimulation. Altered RI potentially underlie the leg motor control impairment after stroke, such as foot drop and development of compensatory strategies. Characterizing the stroke-impaired RI allows future development of targeted strategies to restore this neural function, thereby improving post-stroke gait.
Stroke
Addressing experimental design challenges to investigate stroke-related deficits in the preparation of shoulder movement
Creighton University, Omaha, USA
Stroke
Optimization of a Protocol for Temporary Deafferentation and Proof-of-Concept of Effectiveness for Upper Limb Rehabilitation
University of Texas Rio Grande Valley, Edinburg, USA
Temporary deafferentation (TD) is a rehabilitation technique that uses short-term anesthesia, to inactivate sensation pathways from stronger muscles so that the brain releases inhibition that was placed on weaker muscles, thereby strengthening them. Unfortunately, there are many methods to implement TD, and an optimized protocol for TD has yet to be fully developed. Here, we sought to optimize a protocol for TD to ultimately use as a rehabilitation tool in subjects with spinal cord injury (SCI). Then, as a proof-of-concept, we evaluated hand dexterity and muscle strength to see if there was an improvement after a single session of TD. We hypothesized that the triceps would show a gain in strength, with minimal changes in hand dexterity. We also aimed to determine if our optimized protocol impacted limb electromyography (EMG) after one intervention. For the optimization of our protocol, lidocaine cream (5%) was applied to the right biceps, and sensation was assessed every 15 minutes using von Frey monofilaments. Sensation was assessed for 75 minutes. Percent sensitivity was assessed across time. During proof-of-concept, baseline dexterity, electromyography (EMG) recordings, and strength of biceps and triceps were assessed using the nine-hole peg test, EMG device, and a hand-held dynamometer. Lidocaine cream was applied to the right biceps of 20 healthy volunteers and was removed after 50 minutes. Thirty minutes of exercises to activate their triceps were then performed, and measurements of dexterity and biceps and triceps strength were again recorded. The root-mean-square of the EMG signal was analyzed utilizing LabChart and statistics were evaluated in R and SPSS. Box-Cox transformation was conducted on the dependent variable. We found that peak deafferentation was achieved 50 minutes after lidocaine application. Hence, we conducted a proof-of-concept study combining therapy exercises to activate the triceps after 50 minutes of TD to identify possible effectiveness for upper limb rehabilitation. Our preliminary data suggest that triceps strength is enhanced after one session of TD (p=0.02). Moreover, a single session of TD suggested that the volunteer’s triceps muscle excitability decreased significantly F (1, 329) = 7.66, p < .01. Our results suggest that one session of TD can improve dexterity and biceps strength in healthy subjects. Our analysis infers that five out of the twenty volunteers improved and fifteen did not. The improvement within the triceps muscle excitability varies among the volunteers, X2 = 352, p <.001, and their respective characteristics: BMI, age, fat percentage, and arm width, p < .001, Male gender5, p < .02. X2 = 179.191, 3342,3,4, 5.705. Our protocol will be further refined to determine the effect of numbing cream considering demographics such as BMI, age, body-fat percentage, arm width, and gender to determine the amount of numbing cream administered to a patient.
Spinal Cord Injury (SCI)
Aging-related effects on reference frame utilization during spatial navigation in a novel virtual reality environment
1Georgia Institute of Technology, Atlanta, USA. 2Emory University, Atlanta, USA
Other
Genetic variation in the dopamine system impacts learning response to positive social comparative feedback
University of South Carolina, Columbia, USA
Positive social comparative feedback, or feedback that indicates to the learner that they are performing better than others, is hypothesized to trigger a dopaminergic response that benefits motor learning. Normal genetic variations can result in relatively higher endogenous dopamine transmission versus relatively lower endogenous dopamine transmission, which may impact a learner’s response to a feedback condition that is targeted at the dopamine system. However, no studies have investigated this relationship. The purpose of this secondary data analysis was to examine the interaction between feedback type (response time feedback versus response time plus positive social comparative feedback) and dopamine system genotype on motor sequence learning. Fifty-two individuals (mean age 26.1 ± 4.9; 38 females) practiced a joystick-based motor sequence over two days: Day 1 acquisition and Day 2 retention. Participants were divided into two feedback groups: RT ONLY and RT+POS. The RT ONLY group received feedback about their actual response time to complete the sequence, while the RT+POS group received feedback about their actual response time plus an indication that they were faster than others (i.e., positive social comparative feedback). On Day 2, participants provided saliva samples to allow for genotyping of DRD1 (rs4532), DRD2 (rs1800497), DRD3 (rs6280), and COMT (rs4680). Based on established physiologic impacts of genetic variations on dopamine receptors, each genotype was given a score from 0-2, where higher values indicated higher endogenous dopamine transmission. A summary score (max=8) was generated, and participants were categorized into a low (1-4) or high (5-8) endogenous dopamine transmission group. General linear models with factors for feedback group (RT ONLY; RT+POS), genotype (high; low), and a covariate for State Anxiety (different between groups, p = 0.042) were used to investigate interactions between feedback and genotype group on motor sequences learning (change in response time for a repeated sequence from beginning of acquisition on Day 1 to a retention test on Day 2). We identified a significant genotype by feedback group interaction (p = 0.011). This interaction showed that the high dopamine group benefitted from positive social comparative feedback in a similar manner as response time only (RT ONLY improved by 1.04 ± 0.85 s; RT+POS improved by 1.52 ± 1.11; p = 0.266) while the low dopamine group did not show the same benefit from this kind of feedback (RT ONLY improved by 1.98 ± 0.91s; RT+POS improved by 1.11 ± 0.060; p = 0.008). Positive social comparative feedback may be more beneficial for those with relatively high endogenous dopamine expression. Overall, these results suggest that practice manipulations that rely on a specific neurotransmitter pathway to induce effects may be influenced by transmitter-specific genetic variations. Genotype might be an important factor to consider when implementing strategies to promote specific learning processes.
Motor Rehabilitation
Spasticity can be potentially treated using myoelectrically controlled arm orthosis in chronic stroke
1Cleveland Functional Electrical Stimulation Center, Cleveland, USA. 2Case Western Reserve University, Cleveland, USA. 3University of Washington School of Medicine, Seattle, USA
Stroke
Better Late than Never: Acute Occupational Therapy rehabilitation for Spinal Cord Injury in Low-and-Middle-Income Countries – A case report
1University of Southern California, Los Angeles, USA. 2Christian Medical College, Vellore, India
Spinal Cord Injury (SCI)
Feasibility and preliminary effects of a novel game-based biofeedback interface for stroke gait retraining
1Emory University, Atlanta, USA. 2HiRez Studio, Atlanta, USA. 3Center for Visual and Neurocognitive Rehabilitation Atlanta VA, Atlanta, USA
Reduced paretic propulsion during terminal stance is an important post-stroke gait deficit that can negatively impact gait speed, inter-limb symmetry, and walking function. There is a paucity of gait training approaches that target specific deficits such as paretic propulsion, focus practice preferentially on the paretic leg, and capitalize on motor learning principles to optimize walking. Real-time biofeedback has been shown to be a promising gait retraining strategy. Preliminary data from our laboratory demonstrate that AGRF biofeedback training results in improved paretic leg propulsion and other biomechanical variables, without concomitant changes in the non-paretic leg, thus improving inter-limb symmetry post-stroke. However, our previous gait biofeedback interface was a simple, 2-dimensional, non-engaging display. The objective of this study was to evaluate the feasibility and preliminary effects of a more engaging, game-based system specifically designed for post-stroke AGRF biofeedback gait training. Furthermore, incorporating a user-centered approach, we obtained iterative feedback from key stakeholders (rehabilitation clients and healthcare providers). We hypothesized that post-stroke individuals will demonstrate greater engagement and comparable short-term improvements in gait biomechanics during walking bouts involving biofeedback when training incorporates our novel game-based versus a conventional interface. As part of this study, we collected data on 10 able bodied controls (6 females; age: 27.6± 4.01) and 6 stroke survivors (3 females; age: 65.5± 5.54 years; 5.17 ± 3.66 years since stroke). Study participants completed one session comprising three 4-minute bouts of treadmill walking at self-selected speed with no bio-feedback, as well as walking with 2 different AGRF biofeedback systems in a randomized order – our newly developed, game-based ‘RockWalk’ biofeedback and a conventional, non-gamified biofeedback. Participants were provided scripted instructions regarding the training task (i.e. to increase push-off force with the paretic or right leg) and the mechanics of the biofeedback interface (i.e. how the game works). Individual-specific target AGRF was calculated using the baseline gait trial (15-30% above baseline) to maintain an optimum challenge level for biofeedback. To measure physiological engagement and intensity, rate of perceived exertion, heart rate, and galvanic skin resistance were collected during each walking bout. Gait biomechanics variables were obtained before and after each trial. To obtain user feedback and perceptions, 3 user surveys were implemented after each type of biofeedback. While data analysis is ongoing, our preliminary results related to user feedback provide insights regarding stroke survivors’ perceptions, barriers, and facilitators related to incorporation of sophisticated game-based gait retraining feedback systems. Ongoing and future work will investigate differences in neural mechanisms underlying gamified versus non-gamified biofeedback, comparison of immersive (via virtual reality headsets) versus non-immersive feedback, and the cumulative effects of multiple sessions of game-based biofeedback.
Stroke
Automated Somatosensory Therapy with optional Vagus Nerve Simulation following Nerve Injury
1University of Texas at Dallas, Richardson, USA. 2Vulintus, Lafayette, USA
Somatosensory deficits are common following various types of nerve injury, including hemorrhagic and ischemic stroke, spinal cord injury (SCI), traumatic brain injury (TBI), and peripheral nerve injury (PNI). Stroke is one of the most common causes of chronic disability in the United States, with approximately seven million Americans over the age of 20 currently living with disability after a stroke (Virani et al., 2020). Of those with disabilities, 48% experience physical disability severe enough that they can no longer perform activities of daily living (ADL) independently (Feigenson et al., 1977). Despite the high prevalence of sensory loss in this population, most rehabilitation regimens for disability after stroke focus on specific motor skills rather than tactile therapy (Kaur et al., 2012). Chronic stroke patients have been found to complete a therapy dosage of less than 60 upper limb repetitions per physical therapy visit (Lang et al., 2009), when the recommended dosage for significant positive effects of patient outcomes is closer to 300-500 repetitions per day, repeated several days per week (Lang, et. al.2007, Lang, et al., 2009, Lohse et. al. 2014, Lang, et. al. 2015, Lang, et. al. 2017). The primary focus of this project is to develop and test a novel automated stereognosis testing and training system for use in clinic or at home which can be readily individualized to the patient’s needs in an effort to increase the possible therapy dosage for patients with somatosensory deficits.
Nine participants with chronic brain injury (BI) and self-reported somatosensory deficits on the palmar surface spent an hour using the ReTrieve automated stereognosis system in a clinical setting. Three of the nine participants used the system at home for a week or more to evaluate the potential for ReTrieve to be deployed as an at-home rehabilitation tool for extended use in conjunction with traditional physical therapy. Four participants with spinal cord injury (SCI) used the ReTrieve system during their physical therapy visits for a total of 16 weeks (8 weeks per block) in a crossover study where vagus nerve simulation (VNS) was paired with various automated therapies. At least 8 weeks of therapy was paired with VNS for each participant.
All nine BI participants found the tasks to be engaging and challenging. Errors in object identification were common and yet, all participants were able to identify objects based on object length, shape, texture, or weight at well above chance performance. BI participants who used the system at home completed an average of 38 minutes of automated stereognosis therapy per day. Statistical analysis for both BI and SCI participant data is in progress.
Sensory Rehabilitation
HD-tDCS combined with MusicGlove Gaming Exercises can improve Hand Dexterity in Individuals with Traumatic Brain Injury
1Kessler Foundation, West Orange, USA. 2Rutgers New Jersey Medical School, Newark, USA. 3Biofourmis, Boston, USA
Motor Rehabilitation
Spatial-Motor Training Approaches to Improve Post-Stroke Spatial Neglect
1Emory University, Neuroscience Graduate program, Atlanta, USA. 2Department of Neurology, Atlanta, USA. 3Emory University School of Medicine, Department of Physical Therapy, Atlanta, USA
Aiming spatial neglect (SN) is a dysfunction in execution of motor–intentional behavior that can develop after a right-brain stroke. SN adversely impairs functional mobility resulting in poor rehabilitation outcomes and lower quality of life. Prism adaptation therapy (PAT) is established as an effective treatment to reduce pathological Aiming SN and improve function in people post-stroke. Due to its positive effects on corticomotor excitability, non-invasive sensorimotor electrical stimulation has been used an adjunct to motor training to enhance motor learning. However, the feasibility and interactive effects of electrical stimulation applied in conjunction with PAT on SN are not well-studied. The long-term goal of our research is to evaluate the feasibility and short-term effects of novel spatial-motor treatment strategies that combine PAT with somatosensory stimulation. Here, as a first step towards our goal, our aims were to compare the effects of PAT with versus without electrical stimulation on corticomotor excitability, and the associations of these neurophysiologic effects with the magnitude of PAT-induced sensori-motor adaptation. This study will evaluate the effects of a single session of PAT training with versus without somatosensory stimulation in able-bodied individuals and individuals with right hemisphere stroke (40-90 years, >3 months with SN). Before and after PAT, SN will be evaluated using the computerized line bisection task, and corticomotor excitability will be evaluated using the amplitude of motor evoked potentials (MEPs) elicited by transcranial magnetic stimulation delivered to M1. We hypothesize that PAT with electrical stimulation will induce larger increases in MEP amplitude, and the magnitude of PAT after-effects will be correlated with training-induced change in MEPs. The able-bodied control data will provide important normative references, and mechanistic insights into the effects of PAT. We will present preliminary data on able-bodied participants. Upon completion, our study will generate data supporting novel motor spatial retraining and combinatorial approaches to enhance the rehabilitation of SN and motor function post-stroke.
Stroke
The evolving paradigm of Constraint-Induced Movement Therapy: New findings and conceptual challenges about constraint and neuroplasticity
1Virginia Tech, Roanoke, USA. 2University of Virginia, Charlottesville, USA. 3Ohio State University, Columbus, USA. 4Columbia, New York, USA
Constraint-Induced Movement Therapy (CIMT) began almost three decades ago as a novel therapeutic approach for adults with hemiparesis secondary to stroke and has repeatedly demonstrated efficacy in helping improve outcomes. The research application for children with hemiparetic cerebral palsy (HCP), due to stroke and other causes, began soon thereafter. Today, pediatric CIMT is strongly endorsed via independent, comprehensive research reviews as being efficacious (cf. DeLuca, Ramey & Wolf, 2021). The data related to CIMT effects spans the translational spectrum and adaptations of the treatment protocols and application to different patient populations are actively underway. Examples include the StrokeNet I-ACQUIRE (NCT03910075) a Phase 3 trial testing CIMT for infants and toddlers with perinatal stroke, systematically comparing two different dosage levels; and TRANSPORT 2 (NCT03826030), a Phase 2 trial testing neuromodulation combined with CIMT for adults post-stroke. Two recently completed NIH-funded Phase 2 trials were designed as multisite comparative efficacy trials designed to assess the independent and combined effects of different treatment protocols, varying in dosage and types of constraint (or no constraint) used to treat infants and children with HCP. This presentation will focus on the unexpected and now replicated finding that the constraint of a patient’s unimpaired or less impaired upper extremity produces benefits to constrained side, as well as to the paretic upper extremity (i.e., the primary target of the CIMT treatment). This finding strongly refutes the continued, although not empirically substantiated, caution expressed by some clinicians and parents that the constraint itself could be harmful to the child’s “better side” and could be unduly stressful to the child and family. Just as importantly, this findings leads to provocative questions about potential mediating mechanisms that account for long-term benefits of pediatric CIMT at relatively high dosages – at least 3 hours of therapist implemented CIMT for 5 days a week over 4 consecutive weeks (e.g., Ramey et al, 2021). The presenters will invite discussion about a model that nominates how constraint, when combined with high-dosage treatment, contributes to short- and long-term benefits in CNS changes and as well as “spillover” or multi-domain effects that go beyond the primary outcomes related to use of the paretic upper extremity.
Motor Rehabilitation
Remote Ischemic Conditioning Improves Muscle Strength and Gait Kinematics in Children with Cerebral Palsy
1East Carolina University, Greenville, USA. 2Seattle Childrens Hospital, Seattle, USA
Motor Rehabilitation
Differences in coping styles and perceptions about stroke impairment and stress may correlate with quality of life: An analysis of data from couples in the chronic stage after stroke
1Rosalind Franklin University, North Chicago, USA. 2Captain James A. Lovell Federal Health Care Center, North Chicago, USA. 3Universite de Montreal, Quebec, Canada
Stroke
Ideomotor Apraxia modulates the relationship between functional independence and upper extremity impairment (contralesional and ipsilesional) in chronic stroke survivors with severe paresis
1Penn State College of Medicine, Hershey, USA. 2Pennsylvania State University, University Park, USA. 3University of Southern California, Los Angeles, USA
We previously demonstrated significant ipsilesional functional deficits in chronic stroke patients with a wide range of contralesional impairment severity. We found patients with the most severe contralesional arm impairment had the most severe ipsilesional arm deficits. Previous studies have shown apraxia can potentiate such ipsilesional deficits and is associated with increased reliance on others for ADLs. We now seek to explore this in more detail by considering the relationships between ipsilesional arm impairment, contralesional arm impairment, and functional independence in a group of stroke participants with severe contralesional arm impairment. We assessed the relationship between functional independence (Barthel Index), ipsilesional arm function (grip strength, Jebsen-Taylor Hand Function Test [JHFT]), contralesional arm impairment (Upper-Extremity Fugl-Meyer Assessment [FMA]) in two groups (apraxic, non-apraxic) of participants in the chronic phase post unilateral stroke. Participants with and without apraxia showed significant ipsilesional deficits compared to the previously reported performance of healthy controls. However, participants with apraxia showed a stronger dependence of their functional independence deficits (Barthel Index) on both contralesional impairment (FMA) and on ipsilesional impairments, as measured by grip strength and JHFT. Given that both patients with and without apraxia had substantial ipsilateral and contralateral deficits, the finding that these deficits predict functional independence most strongly in apraxic participants suggests that patients without apraxia may be able to compensate better for their upper-extremity motor deficits. However, further research is warranted with a larger sample of patients with apraxia to explore these findings further.
Stroke
Clinical Application of Vagus Nerve Stimulation Paired with Task Practice for Individuals with Chronic Stroke: Dosage Optimization, Participant Selection, and Training Task Preference
1Emory University School of Medicine, Atlanta, USA. 2Emory University School of Medic, Atlanta, USA
Stroke