Are Kinematic Assessments from a Planar Reaching Task Better Than Standardized Clinical Assessments at Predicting Functional Independence in Chronic Stroke Survivors?

Abstract

Kinematic assessments provide a quantitative evaluation of movement outcomes in chronic stroke survivors. However, it is unclear whether these assessments provide an added benefit to standardized clinical assessments when evaluating functional independence. We hypothesized that kinematic assessments of ipsilesional arm motor and cognitive performance would be better at predicting functional independence compared to their standardized clinical assessment counterparts. We recruited 21 chronic stroke survivors with severe hemiparesis to complete 2 clinical assessments (Jebsen-Taylor Hand Function Test, Grooved Pegboard Test) and 2 kinematic assessments on the Kinereach motion tracking system (a simple reaching task and a cognitively challenging reaching task). We found moderate-to-weak correlations between the Functional Independence Measure (FIM) and each of the kinematic outcomes. The clinical assessments had weak correlations with the FIM. Thus, kinematic assessments provided no significant advantage over clinical assessments in predicting functional independence.

Keywords

motor control movement upper limb ipsilesional FIM

Introduction

Stroke is one of the leading causes of long-term disability and loss of functional independence in the United States (Katan & Luft, 2018). There is evidence of both contralesional and ipsilesional arm motor impairments that may affect functional independence and quality of life of stroke survivors. Therefore, understanding the magnitude of these impairments may serve as a guide for choosing effective therapeutic interventions. Although the majority of research has focused on studying motor deficits in the contralesional arm (Alt Murphy et al., 2013; Carpinella et al., 2020; de Niet et al., 2007), there are many studies that have provided evidence of motor deficits in the ipsilesional arm as well (Jayasinghe, Good, et al., 2020; Kitsos et al., 2013; Maenza et al., 2020; Sainburg & Schaefer, 2004; Smith et al., 2023; Subramaniam et al., 2019). These ipsilesional arm motor deficits may limit an individual's ability to compensate for contralesional arm deficits, resulting in an overall decline in functional independence (Jayasinghe, Good, et al., 2020; Laufer et al., 2001; Subramaniam et al., 2019; Wetter et al., 2005). The side of brain damage may further influence functional outcomes due to hemispheric specialization of motor control. Previous studies have shown that the left hemisphere is specialized for predictive control of movement trajectories, while the right hemisphere contributes more to impedance control and final position accuracy (Sainburg, 2014). Emerging evidence suggests that this hemispheric specialization extends to cognitive functions as well, with the right hemisphere shown to be responsible for emotional recognition, state of mind, etc., and right hemisphere damage particularly impacting visuospatial processing during motor tasks (Bernard et al., 2018; Martín-Monzón et al., 2024).

Clinicians use standardized clinical assessments to make judgments regarding upper limb function and overall functional independence. These assessments have been widely implemented and tested across different age groups, conditions, disease states, and sexes (Jette et al., 2009; Poole, 2011; Santisteban et al., 2016). Thus, there is normative data one can refer to when making decisions related to an individual patient's functional capabilities during rehabilitation. When choosing which standardized clinical assessment to administer, there are several factors that the clinician must consider, including: (1) patient's level of impairment; (2) corresponding statistics of an assessment (e.g., sensitivity, specificity, reliability, validity, and normative data); (3) time and cost associated with conducting the test; (4) familiarity and comfort with administering the test and interpreting the results (Jette et al., 2009; Thompson-Butel et al., 2015). Although it seems sensible to continue using such clinical assessments to evaluate motor function, there are several problems with their use, including: (1) floor and ceiling effects, whereby most participants score near either the minimum or maximum, making it difficult to detect differences between individuals (Thompson-Butel et al., 2015); (2) use of an ordinal score scale versus an interval or ratio scale (i.e., ordinal data does not have a clearly defined interval between values which limits the ability to conduct statistical analyses and draw meaningful conclusions from the results) (Santisteban et al., 2016); (3) variability in inter-rater reliability due to subjectivity (e.g., one examiner might be generally more lenient in assigning scores compared to the other); (4) potential for bias in assigning a score (for example, if the patient would be only considered for therapy if they meet a predetermined cut-off score, then this knowledge might bias the examiner).

Researchers who study movement address limitations of clinical assessments by utilizing more objective and higher-resolution data from kinematics, electromyography, eye-tracking, robotics, etc. (Alt Murphy et al., 2013; Cheng et al., 2023; Dukelow et al., 2010; Marchesi et al., 2021). In neurological populations, the measures derived from these techniques allow for more objective comparisons with neurotypical individuals and can also allow clinicians to categorize impairments more definitively. For example, a study conducted using the Kinarm robot showed the benefits of assessing movement on a continuous data scale rather than on an ordinal scale, including the ability to differentiate between minor changes in deficits over time (Dukelow et al., 2010). However, there are issues with these “higher resolution” assessments as well. First, interpretation of the data derived from these assessments requires understanding the specific experimental context. For example, a change in endpoint error from 4 cm to 2 cm during a center-out reaching task is not comparable to the same change if the starting location of the limb were changed. Second, clinical implications of changes in motor performance are not easily understood since there is no standardization of these experimental paradigms (Dukelow et al., 2010); hence, the results derived from an experiment conducted using one system may differ from those conducted on a different system. Third, although the collection of high-resolution data can be conducted relatively objectively, there is a certain amount of subjectivity when determining filtering thresholds during the processing of movement trajectories even when following strict analytical procedures, and especially if such processing cannot be automated and requires human input.

Researchers have sought to understand which set of standardized clinical assessments best provide the information necessary to predict functional independence and upper limb function in stroke survivors. Studies have been inconclusive in finding one best assessment and instead suggest the value of implementing several assessments to ensure the most clinically accurate picture of the patient's level of function and independence (Bushnell et al., 2015; Santisteban et al., 2016; Thompson-Butel et al., 2015). Many studies that have examined the effectiveness of kinematic assessments in predicting upper limb motor function in stroke survivors have correlated kinematic outcomes with scores on standardized clinical assessments (Alt Murphy et al., 2013; Carpinella et al., 2020; Marchesi et al., 2021). We used this method to examine whether motor and cognitive deficits predicted functional independence differently depending on the side of brain damage, and we found that while both left and right hemisphere stroke damage results in motor deficits, only those with left hemisphere damage produced a strong correlation between these deficits and their functional independence level (Jayasinghe, Good, et al., 2020). However, there are still some important gaps in the stroke literature that need to be resolved, including whether: (1) higher-resolution assessments of motor function are better at predicting functional independence compared to lower-resolution assessments; (2) the side of brain damage affects which type of assessments is better at predicting functional independence; (3) evaluation of ipsilesional arm deficits is better with higher-resolution assessments since they may be more subtle than those in the contralesional arm. Prior work has established the relationship between contralesional arm impairment and overall functional independence (Coupar et al., 2012; Rabadi & Rabadi, 2006), but the role of the ipsilesional arm motor and cognitive deficits in achieving functional independence remains unresolved.

To assess the effect of ipsilesional arm motor deficits on functional independence, we collected data from standardized clinical assessments of upper limb function (i.e., the lower-resolution assessments) and from kinematics data in two reaching tasks completed on a virtual reality motion capture system (i.e., the higher-resolution assessments). We focused specifically on ipsilesional arm performance because: (1) severe hemiparesis leads to greater reliance on the ipsilesional arm for daily activities (Kitsos et al., 2013; Maenza et al., 2020; Smith et al., 2023), (2) subtle deficits in the ipsilesional arm may be missed by clinical assessments but captured by kinematics analyses, and (3) prior work in our laboratory shows that ipsilesional arm motor impairments correlated with functional independence, particularly after left hemisphere damage. We hypothesized that the higher-resolution assessments would provide a stronger correlation with functional independence compared to the lower-resolution assessments. We also examined whether the side of brain damage influenced the relationship between kinematic assessments and functional independence. Based on prior evidence of hemispheric specialization (Jayasinghe, Sarlegna, et al., 2020; Sainburg, 2014), we constructed a preliminary hypothesis that individuals with left hemisphere damage would exhibit a stronger correlation between functional independence and kinematic assessments of motor coordination, whereas individuals with right hemisphere damage would show a stronger correlation between functional independence and kinematic assessments of cognitive aspects of movement. This laterality-specific approach could inform more targeted rehabilitation strategies. For instance, therapists may want to emphasize movement accuracy training after left hemisphere damage versus cognitive-motor integration after right hemisphere damage. We had a third (exploratory) hypothesis—the side of brain damage influenced which type of assessment (i.e., clinical vs. kinematic) would be better at predicting overall functional independence. Through this research, we aim to understand how we can better predict functional independence to further guide therapeutic interventions to optimize and improve the quality of life for stroke survivors.

Methods

Participants

This experiment is a part of a larger ongoing study in our lab. We recruited 21 chronic stroke survivors (8 left hemisphere damage, 13 right hemisphere damage; 11 males, 10 females; age 57 ± 2.51 years; right hand dominant pre-stroke; >3 months post stroke) with severe paresis in their contralesional arm. Severe paresis was classified as a score <29 (Woytowicz et al., 2017) on the upper-extremity portion of the Fugl-Meyer assessment (FMA-UE) (Fugl-Meyer et al., 1975). Hand dominance was confirmed via the Edinburgh Handedness Inventory (Oldfield, 1971). We used the Grooved Pegboard Test (GPT) (Lafayette Instruments, Lafayette, IN) as a measure of general executive and cognitive function (Ashendorf et al., 2009; Bezdicek et al., 2014; Tolle et al., 2020). Exclusion criteria included the presence of bilateral lesions, non-stroke related neurological conditions, and movement-limiting pain. We obtained written informed consent from all participants prior to beginning the study. All study procedures were approved by the University of Minnesota's Institutional Review Board.

Experimental Design

Clinical assessments. Table 1 provides the clinical assessment scores obtained from each study participant. We used the Jebsen-Taylor Hand Function Test (JTHFT) and Grooved Pegboard Test (GPT) to determine ipsilesional arm motor and cognitive function, respectively. These were deemed the lower-resolution assessments in our study. We assessed overall function using the Functional Independence Measure (FIM). The JTHFT is used to assess hand function by engaging participants in tasks that simulate activities of daily living, such as writing, page turning, picking up objects, simulated feeding, stacking checkers, and picking up light and heavy cans (Jebsen et al., 1969). We included the JTHFT score without the writing component because writing with the nondominant hand is known to take longer than with the dominant hand and can result in a higher JTHFT score solely due to the writing score. The GPT is a test of dexterity and hand coordination in which the individual must quickly place 25 pegs one-by-one in grooves on a board and then remove them quickly, one-by-one (Kløve, 1963; Matthews & Kløve, 1964). Studies have provided support for the utility of the GPT as a measure of executive function deficits and general cognitive decline (Ashendorf et al., 2009; Bezdicek et al., 2014; Tolle et al., 2020); thus, we utilized the GPT in this study as a measure of deficits related to cognitive aspects of movement. We used the self-care portion of the FIM (score out of 42) to assess the level of independence in completing a few everyday tasks, such as eating, grooming, and dressing (Kidd et al., 1995). Higher scores on the FIM suggest a higher level of independence.

Table 1.

Demographics and Clinical Assessment Scores of Study Participants. Chronicity (time post stroke); Lesion side: RHD (Right Hemisphere Damage), LHD (Left Hemisphere Damage; FIM: Functional Independence Measure; FMA-UE: Upper Extremity portion of the Fugl-Meyer Assessment conducted on the contralesional arm; GPT: Grooved Pegboard Test (total time taken for the ipsilesional hand to place the pegs on the board and then remove them); EHI: Edinburgh Handedness Inventory; JTHFT: Jebsen-Taylor Hand Function Test (ipsilesional hand total score without the writing component).

ID	Age (years)	Sex (M/F)	Chronicity (years)	Lesion Side (LHD/RHD)	FIM (/42)	FMA-UE (/66)	GPT (s)	EHI (%)	JTHFT (s)
S1	62	M	8.1	RHD	32	17	106.96	100	44.45
S2	59	M	13.7	RHD	38	13	95.06	100	29.53
S3	35	M	10.7	RHD	32	24	100.85	100	34.20
S4	49	M	1.8	RHD	37	12	98.56	100	49.13
S5	55	M	2.0	RHD	29	28	115.53	100	39.99
S6	78	M	2.0	RHD	35	23	104.90	100	35.93
S7	55	F	7.6	RHD	30	13	102.28	100	33.74
S8	66	F	15.1	RHD	32	11	109.34	89.5	32.06
S9	71	F	1.3	RHD	40	14	97.59	100	35.86
S10	55	F	10.2	RHD	28	4	125.46	89.5	40.60
S11	56	F	0.5	RHD	18	4	96.18	90	35.43
S12	68	M	8.7	RHD	37	22	116.76	90	37.81
S13	39	F	0.4	RHD	38	9	80.13	90	41.67
S14	59	F	12.0	LHD	35	15	111.65	90	44.36
S15	45	F	13.3	LHD	33	17	113.22	100	43.23
S16	57	F	2.7	LHD	34	5	83.82	90	32.14
S17	58	M	2.3	LHD	36	13	92.56	100	34.45
S18	44	M	6.9	LHD	36	14	114.87	100	41.78
S19	77	F	6.1	LHD	34	10	119.09	90	35.31
S20	64	M	5.2	LHD	35	16	180.16	100	42.34
S21	46	M	2.0	LHD	35	14	99.75	90	43.71

Kinematics. Assessment of kinematics data provided the higher-resolution measures of ipsilesional arm motor and cognitive function. We used the Kinereach virtual reality motion capture system to obtain the kinematics data (Figure 1a). Participants were seated in front of a table with a horizontally orientated mirror. An inverted HD monitor above the mirror projected the task onto the mirror, which allowed the tasks to be shown in the same plane as the ipsilesional arm. We attached two sensors (trakStar^®; NDI) to the hand and upper arm to record movement at 116 Hz. The hand was braced and then placed on an air sled that reduced the effects of friction and gravity during arm movement on the glass tabletop. Each participant completed two tasks on the Kinereach system—a simple reaching task and a cognitively challenging reaching task (Jayasinghe, 2025). In the simple reaching task (Figure 1b), participants were instructed to move a cursor (representing their hand position in real-time) from the specified start position to the target as quickly as possible within 1 s. There were 99 trials in this task, and the target (3.5 cm diameter) could appear pseudorandomly in one of three locations 17 cm from the start position. The cognitively challenging reaching task (Figure 1c) involved participants having to quickly commit to memory a set of pictorial instructions before identifying and reaching towards the correct target in an array of objects within 3 s. The target was 15 cm from the start circle and was either 3 cm or 4.5 cm in diameter. The cognitive load gradually increased over the course of the experiment (170 trials) as a combination of increased complexity of the target array as well as the number of instructions to remember. Both tasks were self-paced, such that participants could take breaks in-between trials. The simple reaching task is a purely motor task and was used to derive measures of ipsilesional arm motor deficits while the cognitively challenging reaching task was used to derive a kinematic measure of deficits related to cognitive aspects of movement.

Figure 1.

Schematic of the experimental setup and task. (a) The Kinereach setup. Participants were seated in front of a table with a horizontally oriented mirror that displayed the task and blocked view of the hands. Sensors were placed on the ipsilesional hand and upper arm to collect position and orientation data of the upper limb. (b) The simple reaching task. In this task, a trial consisted of the target appearing above the start circle in one of 3 locations. Participants were instructed to quickly move the cursor towards the target within a 1 s period. (c) The cognitively challenging reaching task. In this task, a trial consisted of 2 screens—Screen 1 appeared for 2 s with a set of pictorial cues to commit to memory, then Screen 2 would appear for 3 s with an object array from which the participant needed to locate and reach for the correct target. The example here is representative of the highest level of cognitive load that would be provided.

Data Analysis

We processed and analyzed all kinematics data using custom programs developed in IgorPro (version 9, WaveMetrics, Portland, OR). We examined the following outcome measures: reaction time, deviation from linearity and end position error. Reaction time was calculated from the cognitively challenging reaching task as the time between the appearance of the object array and onset of movement, and it provided a measure of cognitive function during a motor task. Deviation from linearity was calculated from the simple reaching task and determined by dividing the hand path's minor axis by its major axis (the major axis was defined as the largest distance between any two points on the hand path, and the minor axis was the largest distance that is perpendicular to the major axis). End position error was also calculated from the simple reaching task and defined as the distance between the target and cursor at the end of trial. These latter two variables provided a measure of motor function.

All clinical assessment scores and statistical analyses were analyzed using JMP Pro (Version 17, SAS Institute, Cary, NC). Normality and variance of our data was tested using the Shapiro-Wilk and Levene test, respectively. We applied a transformation to the data when necessary and used nonparametric tests only for cases where normal distributions could not be achieved. We used either a pooled t test or the Mann-Whitney U test, which is the nonparametric alternative of the pooled t test, to determine group differences in left hemisphere damage (LHD) vs. right brain damage (RHD), for each of the following variables: FIM, JTHFT, GPT, reaction time, deviation from linearity and end position error. We performed linear correlations to test our hypothesis. We used Pearsons's r (two-tailed significance) to assess the linear relationship between FIM and each of the kinematic and clinical assessments of ipsilesional arm motor function. We then used cocor (Diedenhofen & Musch, 2015), an online statistical analysis tool, to compare the correlation coefficients. We used the output of Pearson and Filon's test from the cocor package to report those results. We used a Type I error rate of 0.05.

Results

Participants Characteristics

We found no statistically significant differences between the left and right hemisphere damage groups with respect to each of the kinematic outcome measures: reaction time (t = 0.001, p = 0.999), deviation from linearity (t = −1.10, p = 0.286), end position error (t = 1.70, p = 0.106). Similarly, we found no significant differences between groups for each of the clinical assessment scores: JTHFT (t = −0.83, p = 0.414), GPT (t = −1.06, p = 0.304), FIM (U = 0.51, p = 0.610).

Kinematic Assessments are not Better at Predicting Functional Independence Compared to Clinical Assessments in Chronic Stroke Survivors

Figure 2 shows our examination of the relationship between the Functional Independence Measure (FIM) score and each of the kinematic measures (reaction time, deviation from linearity and end position error) and each of the clinical assessment scores (JTHFT and GPT). We found that chronic stroke survivors exhibited a moderate linear relation between FIM and reaction time [r₍₂₁₎ = −0.38, p = 0.089] but weak linear relations between each of FIM and deviation from linearity [r₍₂₁₎ = 0.23, p = 0.311] and FIM and end position error [r₍₂₁₎ = −0.24, p = 0.303], respectively. Thus, the correlation between functional independence and cognitive function was stronger than that with motor function, at least with the kinematic assessments. With respect to the clinical assessments, we found weak linear relations between FIM and GPT [r₍₂₁₎ = −0.06, p = 0.813] and FIM and JTHFT [r₍₂₁₎ = 0.11, p = 0.627]. This suggests that neither of the motor or cognitive clinical assessments were able to predict functional independence well. The cocor analyses revealed no significant differences in correlation strength between kinematic assessments and clinical assessments (Table 2), indicating that kinematic measures did not outperform clinical assessments in predicting functional independence.

Figure 2.

Linear relations between functional independence and ipsilesional arm kinematic and clinical assessments of motor and cognitive function in chronic stroke survivors. The FIM score was moderately corelated with reaction time (n = 21) and weakly correlated with deviation from linearity (n = 21), end position error (n = 21), grooved pegboard test (n = 21) and Jebsen-Taylor Hand Function Test (n = 21). Linear fit and Pearson’s correlation value are shown for each relationship. See Table 2 for statistical analyses.

Table 2.

Comparison of Correlation Strength Between FIM and each of the Kinematic and Clinical Assessments. We used cocor to compare each pair of correlations. FIM: Functional Independence Measure; GPT: Grooved Pegboard Test; JTHFT: Jebsen-Taylor Hand Function Test.

	Kinematic measure	Clinical assessment	Hypothesis	p value
FIM	Reaction time (r₁)	GPT (r₂)
	−0.38	−0.06	r₁ _> r₂	0.904
	Linearity (r₁)	JTHFT (r₂)
	0.23	0.11	r₁ _> r₂	0.332
	End position error (r₁)	JTHFT (r₂)
	−0.24	0.11	r₁ _> r₂	0.881

Functional Independence Correlates with various Kinematic Assessments Differently in RHD Compared to LHD Individuals

Figure 3 shows the relationship between the FIM score and each of the kinematic measures and each of the clinical assessments of ipsilesional arm performance in each group (i.e., LHD and RHD). In terms of kinematic measures, the RHD group exhibited a moderate linear relation between FIM and reaction time [r₍₁₃₎ = −0.43, p = 0.140], and a weak linear relation between each of FIM and deviation from linearity [r₍₁₃₎ = 0.27, p = 0.367] and end position error [r₍₁₃₎ = −0.16, p = 0.595]. The LHD group displayed a weak linear relation between FIM and reaction time [r₍₈₎ = −0.14, p = 0.741], and a strong linear relation between each of FIM and deviation from linearity [r₍₈₎ = −0.57, p = 0.144] and FIM and end position error [r₍₈₎ = −0.67, p = 0.071]. In terms of clinical assessments, the RHD group exhibited a weak linear relation between each of FIM and GPT [r₍₁₃₎ = −0.27, p = 0.382], and FIM and JTHFT [r₍₁₃₎ = 0.09, p = 0.773]. The LHD group also displayed a weak linear relation between FIM and GPT[r₍₈₎ = 0.03, p = 0.939], and FIM and JTHFT[r₍₈₎ = 0.05, p = 0.912]. Thus, the kinematic assessments produced moderate-to-strong correlations with the FIM score in each of the LHD and RHD groups for their respective aspect of control (i.e., motor coordination in LHD and cognitive aspects of motor control in RHD). We tested the strength of these correlations, using cocor, to determine whether the side of brain damage influenced which type of assessment (i.e., clinical vs. kinematic) would be better at predicting functional independence. As seen in Table 3, our hypotheses were not supported.

Figure 3.

Linear relations between functional independence and ipsilesional arm kinematic and clinical assessments of motor and cognitive function in LHD and RHD. The FIM score is moderately correlated with reaction time in RHD (n = 13) and strongly correlated with deviation from linearity and end position error in LHD (n = 8) group. The FIM score is weakly correlated with each of GPT and JTHFT in both groups. Linear fit and Pearson’s correlation value are shown for each relationship. See Table 3 for statistical analyses.

Table 3.

Comparison of Correlation Strength between FIM and each of the Kinematic and Clinical Assessments in each Group (i.e., LHD and RHD). We used cocor to Conduct These Comparisons. FIM: Functional Independence Measure; GPT: Grooved Pegboard Test; JTHFT: Jebsen-Taylor Hand Function Test.

	RHD		Hypothesis	p value		LHD		Hypothesis	p value
	Kinematic measure	Clinical assessment	Hypothesis	p value		Kinematic measure	Clinical assessment	Hypothesis	p value
	Reaction time (r₁)	GPT (r₂)				Reaction time (r₁)	GPT (r₂)
FIM	−0.43	−0.27	r₁ ≠ r₂	0.596	FIM	−0.14	0.03	r₁ ≠ r₂	0.685
	Linearity (r₁)	JTHFT (r₂)				Linearity (r₁)	JTHFT (r₂)
FIM	0.27	0.09	r₁≠ r₂	0.605	FIM	−0.57	0.05	r₁ ≠ r₂	0.112
	End position error (r₁)	JTHFT (r₂)				End position error (r₁)	JTHFT (r₂)
FIM	−0.16	0.09	r₁≠ r₂	0.516	FIM	−0.67	0.05	r₁ ≠ r₂	0.072

Discussion

In this study, we examined whether higher-resolution kinematic assessments of the ipsilesional arm were better than lower-resolution clinical assessments at predicting functional independence in 21 chronic stroke survivors with severe hemiparesis. Contrary to our primary hypothesis, kinematic assessments did not significantly outperform clinical assessments of ipsilesional arm function (i.e., JTHFT and GPT) in predicting functional independence. However, we found lesion side-specific correlations between kinematic assessments and functional independence (i.e., a strong correlation between FIM and deviation from linearity and FIM and end position error in LHD and a moderate correlation between FIM and reaction time in RHD) that may be relevant for clinical implementation. We did not find support for our exploratory hypothesis that the side of brain damage influenced whether a clinical or kinematic assessment was better at predicting FIM scores. Overall, these findings suggest that ipsilesional arm deficits, particularly those detectable via kinematics, may contribute to predicting overall functional independence, in addition to contralesional impairments.

The hemispheric differences observed align with established models of left hemisphere specialization for predictive motor control and right hemisphere specialization in visuospatial processing and impedance control (Jayasinghe, Good, et al., 2020; Sainburg, 2014). We found that individuals with left hemisphere damage (LHD) showed a strong correlation between the FIM score and measures of motor coordination (i.e., end position error and deviation from linearity), while individuals with right hemisphere damage (RHD) showed a moderate correlation with cognitive aspects of motor control (i.e., reaction time). For individuals with left hemisphere damage, higher end position errors may reflect impaired trajectory planning which can directly impact compensatory use of the ipsilesional arm during bilateral tasks, like buttoning a shirt or using a fork and knife to eat. In individuals with right hemisphere damage, slower reaction times may manifest in activities of daily living in the form of delayed reaction to time-sensitive activities, such as reaching for an item falling off the counter. Kinematic assessments may be used to uncover these hemisphere-specific compensatory challenges that may otherwise be overlooked in clinical assessments. These results are supported by previous and ongoing work in our lab that examine the role of lesion laterality on how well cognitive and motor outcomes each predict functional independence and how they may guide therapeutic interventions (Jayasinghe, Good, et al., 2020). In our ongoing work, we have found that linear relations between functional independence measures and kinematic outcomes in each of the LHD and RHD groups provide evidence of a larger role of the right hemisphere in cognitive aspects of motor function and a larger role of the left hemisphere in motor coordination.

Although our results refuted our primary hypothesis of the superiority of kinematic measures compared to clinical assessments, we found varying strengths of correlations between the kinematic measures and the FIM score. We found that the FIM score was moderately correlated with a kinematic measure of cognitive function (i.e., reaction time) but was weakly correlated with its clinical counterpart (i.e., GPT). However, we found weak correlations between the FIM score and both the kinematic measures of motor control (i.e., deviation from linearity, end position error) as well as with the corresponding clinical measure of motor control (i.e., JTHFT). The weaker correlations between the clinical assessments (i.e., JTHFT, GPT) and FIM score suggest that the nature of the tasks outlined in the FIM may have influenced predictive validity. While kinematic measures captured hemisphere-specific deficits in movement control, the FIM assessed global independence in activities of daily living, which often involve compensatory strategies post-stroke (i.e., using assistive devices or the less impaired limb); this is not reflected in the JTHFT or GPT. Standardized assessments of motor function, like the Wolf Motor Function Test or the Action Research Arm Test, which incorporate ADL-like tasks (i.e., gripping objects, lifting cans, touching the back of the head, etc.) may better align with the functional independence tasks outlined in the FIM and should be investigated in future work. Additionally, the FIM's reliance on experimenter-rated performance that is based on the participant's verbal account of behaviors may introduce additional variability unrelated to objective motor and cognitive deficits.

We acknowledge a few limitations that may have contributed to our results lacking strong support for our hypotheses. The small sample size and unequal distribution of LHD vs. RHD participants could have increased Type II error and reduced statistical power, thus, limiting the generalizability of our results. We utilized three kinematic variables that corresponded to cognitive (i.e., reaction time) and motor function (i.e., deviation from linearity, end position error). Although we had a clear rationale for choosing these variables, it is possible that other measures of kinematic performance (such as peak velocity, initial direction error, grip strength, ROM, etc.) may have provided a more comprehensive understanding of cognitive and motor deficits. Further research could explore additional high-resolution methods (i.e., robotic-assisted technology, eye-tracking, electromyography, etc.) for fine motor and cognitive assessments. Another limitation was the inclusion of only right-handed (pre-stroke) stroke survivors in our study, which could have moderated the relationships we tested since hand dominance may have affected the amount of arm use. Although we controlled for pre-stroke handedness, we did not quantify post-stroke changes in hand preference. Previous research has shown that individuals compensating with their original nondominant arm may develop distinct movement strategies than those using their original dominant arm (Lang et al., 2017). We also acknowledge that the clinical assessments used in this study introduced subjectivity in administration, scoring, and interpretation. This could be mitigated by repeated testing over multiple sessions, interrater reliability checks, or ensuring standardized administration via extensive instructor training, even though these approaches have drawbacks (i.e., practice effects, participant burden, examiner bias). Kinematic measures help avoid these issues by automating data collection, but their clinical adoption requires balancing practicality with applicability. The nature of our planar reaching task did not allow for vertical motion and was not able to simulate the complexity of movement in the real-world setting. While this may have affected the strength of correlation between kinematic outcomes and the functional independence measure, we believe that 2D movement helps us understand foundational principles of movement, and by supporting the arm on an air sled in this scenario, it reduced the mechanical effects of friction gravity on the individual's arm.

Although our study employed high-resolution kinematics, the planar reaching paradigm could be adapted in the clinical setting by using low-cost tools, such as towel slides (for learning to reduce path deviation) or tablet applications (for learning to reduce reaction time). The moderate-to-strong correlations between the FIM score and kinematic variables suggest that even simplified versions of these Kinereach tasks could provide meaningful functional insights. For example, a towel slide test quantifying movement time or path deviation might capture similar motor control deficits found in high-resolution kinematics. We acknowledge that the large number of trials in our experiment (99 trials in the simple reaching task and 170 trials in the cognitively challenging reaching task) is not feasible in the clinical setting. Future work should determine the minimum number of trials necessary to obtain reliable kinematic metrics that provide support for these results. This will be essential for evaluating whether the proposed kinematic assessments are practical and efficient in the clinic. Low-cost tools could make ipsilesional arm assessment accessible within resource-limited clinics, while maintaining focus on functionally relevant kinematic components, ultimately supporting more personalized rehabilitation approaches.

Despite the lack of support for our primary hypothesis, we expect that kinematic assessments may complement clinical measures by detecting subtle ipsilesional deficits that may go unnoticed with standardized clinical assessments. For effective clinical implementation, it would be helpful to provide exercises that focus on improving movement linearity and accuracy for those with left brain damage, and exercises that focus on improving reaction time in those with right brain damage. Thus, having the ability to obtain quantitative data to confirm the presence of these subtle deficits would provide valuable information for clinicians when determining rehabilitation strategies. To our knowledge, this is the first study aimed at directly comparing clinical assessments and kinematic assessments to determine which of these would be better predictors of functional independence in chronic stroke survivors. While further research is needed to validate these findings and provide generalizability of our study results, the integration of kinematic assessments in rehabilitation practice holds promise for providing individualized stroke rehabilitation protocols for optimal motor recovery. Future studies should involve longitudinal designs to determine whether performing kinematic assessments early in post-stroke rehabilitation can predict long-term functional outcomes and guide targeted therapeutic interventions.

Footnotes

Acknowledgments

We thank all research participants who contributed their time to this study. We also acknowledge Michele Darger and Lelti Asgedom for their efforts in recruitment and data collection.

ORCID iDs

Pramisha Thapa

Shanie AL Jayasinghe

Ethical Considerations

This study was approved by the University of Minnesota Institutional Review Board (STUDY00015809) on 06/01/2022.

Consent to Participate

All participants provided written informed consent and HIPAA Authorization to participate in this study.

Consent for Publication

No participant identifying information is shared in this publication. All participants were notified that their data may be used in future publications during the informed consent process.

Author Contributions

Study conception and design: SALJ; data collection and data analysis: BA, ZK, PT, SALJ; interpretation of results: BA, ZK, PT, SALJ; manuscript preparation: BA, ZK, PT, SALJ; funding: SALJ

Funding

This work was supported by the National Institutes of Health (R21HD111748; PI: SALJ); the University of Minnesota Foundation Assistant Professor Award (PI: SALJ); the National Institutes of Health C-STAR Collaborative Mentorship Program (P2CHD101899; subcontract PI: SALJ); the National Center for Advancing Translational Sciences (UL1TR002494); New Student Award, University of Minnesota Rehabilitation Science program (awarded to PT).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability

The data associated with the findings of this study are available from the corresponding author upon reasonable request.

References

Alt Murphy

Willén

Sunnerhagen

K. S.

(2013). Responsiveness of upper extremity kinematic measures and clinical improvement during the first three months after stroke. Neurorehabilitation and Neural Repair, 27(9), 844–853. https://doi.org/10.1177/1545968313491008

Ashendorf

Vanderslice-Barr

J. L.

McCaffrey

R. J.

(2009). Motor tests and cognition in healthy older adults. Applied Neuropsychology, 16(3), 171–176. https://doi.org/10.1080/09084280903098562

Bernard

Lemée

J.-M.

Ter Minassian

Menei

(2018). Right hemisphere cognitive functions: From clinical and anatomic bases to brain mapping during awake craniotomy part I: Clinical and functional anatomy. World Neurosurgery, 118, 348–359. https://doi.org/10.1016/j.wneu.2018.05.024

Bezdicek

Nikolai

Hoskovcová

Štochl

Brožová

Dušek

Zárubová

Jech

Ružička

(2014). Grooved pegboard predicates more of cognitive than motor involvement in Parkinson’s disease. Assessment, 21(6), 723–730. https://doi.org/10.1177/1073191114524271

Bushnell

Bettger

J. P.

Cockroft

K. M.

Cramer

S. C.

Edelen

M. O.

Hanley

Katzan

I. L.

Mattke

Nilsen

D. M.

Piquado

, & others . (2015). Chronic stroke outcome measures for motor function intervention trials: Expert panel recommendations. Circulation: Cardiovascular Quality and Outcomes, 8(6_suppl_3), S163–S169. https://doi.org/10.1161/CIRCOUTCOMES.115.002098

Carpinella

Lencioni

Bowman

Bertoni

Turolla

Ferrarin

Jonsdottir

(2020). Effects of robot therapy on upper body kinematics and arm function in persons post stroke: A pilot randomized controlled trial. Journal of Neuroengineering and Rehabilitation, 17(1), 10. https://doi.org/10.1186/s12984-020-0646-1

Cheng

K. Y.

Rehani

Hebert

J. S.

(2023). A scoping review of eye tracking metrics used to assess visuomotor behaviours of upper limb prosthesis users. Journal of NeuroEngineering and Rehabilitation, 20(1), 49. https://doi.org/10.1186/s12984-023-01180-1

Coupar

Pollock

Rowe

Weir

Langhorne

(2012). Predictors of upper limb recovery after stroke: A systematic review and meta-analysis. Clinical Rehabilitation, 26(4), 291–313. https://doi.org/10.1177/0269215511420305

de Niet

Bussmann

J. B.

Ribbers

G. M.

Stam

H. J.

(2007). The stroke upper-limb activity monitor: Its sensitivity to measure hemiplegic upper-limb activity during daily life. Archives of Physical Medicine and Rehabilitation, 88(9), 1121–1126. https://doi.org/10.1016/j.apmr.2007.06.005

10.

Diedenhofen

Musch

(2015). Cocor: A comprehensive solution for the statistical comparison of correlations. PLOS ONE, 10(4), e0121945. https://doi.org/10.1371/journal.pone.0121945

11.

Dukelow

S. P.

Herter

T. M.

Moore

K. D.

Demers

M. J.

Glasgow

J. I.

Bagg

S. D.

Norman

K. E.

Scott

S. H.

(2010). Quantitative assessment of limb position sense following stroke. Neurorehabilitation and Neural Repair, 24(2), 178–187. https://doi.org/10.1177/1545968309345267

12.

Fugl-Meyer

A. R.

Jääskö

Leyman

Olsson

Steglind

(1975). The post-stroke hemiplegic patient. 1. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 7(1), 13–31. https://doi.org/10.2340/1650197771331

13.

Jayasinghe

S. A. L.

Good

Wagstaff

D. A.

Winstein

Sainburg

R. L.

(2020). Motor deficits in the ipsilesional arm of severely paretic stroke survivors correlate with functional independence in left, but not right hemisphere damage. Frontiers in Human Neuroscience, 14, 599220. https://doi.org/10.3389/fnhum.2020.599220

14.

Jayasinghe

S. A. L.

Sarlegna

F. R.

Scheidt

R. A.

Sainburg

R. L.

(2020). The neural foundations of handedness: Insights from a rare case of deafferentation. Journal of Neurophysiology, 124(1), 259–267. https://doi.org/10.1152/jn.00150.2020

15.

Jayasinghe

S. A. L.

(2025). The role of cognitive load on interlimb differences in motor coordination in older adults. Journal of Neurophysiology, 133(1), 60–68. https://doi.org/10.1152/jn.00167.2024

16.

Jebsen

R. H.

Taylor

Trieschmann

R. B.

Trotter

M. J.

Howard

L. A.

(1969). An objective and standardized test of hand function. Archives of Physical Medicine and Rehabilitation, 50(6), 311–319.

17.

Jette

D. U.

Halbert

Iverson

Miceli

Shah

(2009). Use of standardized outcome measures in physical therapist practice: Perceptions and applications. Physical Therapy, 89(2), 125–135. https://doi.org/10.2522/ptj.20080234

18.

Katan

Luft

(2018). Global burden of stroke. Seminars in Neurology, 38(02), 208–211. https://doi.org/10.1055/s-0038-1649503

19.

Kidd

Stewart

Baldry

Johnson

Rossiter

Petruckevitch

Thompson

A. J.

(1995). The functional independence measure: A comparative validity and reliability study. Disability and Rehabilitation, 17(1), 10–14. https://doi.org/10.3109/09638289509166622

20.

Kitsos

G. H.

Hubbard

I. J.

Kitsos

A. R.

Parsons

M. W.

(2013). The ipsilesional upper limb can be affected following stroke. The Scientific World Journal, 2013(1), 684860. https://doi.org/10.1155/2013/684860

21.

Kløve

(1963). Clinical neuropsychology. Medical Clinics of North America, 47(6), 1647–1658. https://doi.org/10.1016/S0025-7125(16)33515-5

22.

Lang

C. E.

Waddell

K. J.

Klaesner

J. W.

Bland

M. D.

(2017). A method for quantifying upper limb performance in daily life using accelerometers. Journal of Visualized Experiments, 122, 55673. https://doi.org/10.3791/55673

23.

Laufer

Gattenio

Parnas

Sinai

Sorek

Dickstein

(2001). Time-related changes in motor performance of the upper extremity ipsilateral to the side of the lesion in stroke survivors. Neurorehabilitation and Neural Repair, 15(3), 167–172. https://doi.org/10.1177/154596830101500303

24.

Maenza

Good

D. C.

Winstein

C. J.

Wagstaff

D. A.

Sainburg

R. L.

(2020). Functional deficits in the less-impaired arm of stroke survivors depend on hemisphere of damage and extent of paretic arm impairment. Neurorehabilitation and Neural Repair, 34(1), 39–50. https://doi.org/10.1177/1545968319875951

25.

Marchesi

Ballardini

Barone

Giannoni

Lentino

De Luca

Casadio

(2021). Modified functional reach test: Upper-body kinematics and muscular activity in chronic stroke survivors. Sensors, 22(1), 230. https://doi.org/10.3390/s22010230

26.

Martín-Monzón

Amores-Carrera

Sabsevitz

Herbet

(2024). Intraoperative mapping of the right hemisphere: A systematic review of protocols that evaluate cognitive and social cognitive functions. Frontiers in Psychology, 15, 1415523. https://doi.org/10.3389/fpsyg.2024.1415523

27.

Matthews

C. G.

Kløve

(1964). Neuropsychological test battery manual. University of Wisconsin Neuropsychology Laboratory.

28.

Oldfield

R. C.

(1971). The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9(1), 97–113. https://doi.org/10.1016/0028-3932(71)90067-4

29.

Poole

J. L.

(2011). Measures of hand function: Arthritis Hand Function Test (AHFT), Australian Canadian Osteoarthritis Hand Index (AUSCAN), Cochin Hand Function Scale, Functional Index for Hand Osteoarthritis (FIHOA), Grip Ability Test (GAT), Jebsen Hand Function Test (JHFT), and Michigan Hand Outcomes Questionnaire (MHQ). Arthritis Care & Research, 63(S11), S189–199. https://doi.org/10.1002/acr.20631

30.

Rabadi

M. H.

Rabadi

F. M.

(2006). Comparison of the action research arm test and the Fugl-Meyer assessment as measures of upper-extremity motor weakness after stroke. Archives of Physical Medicine and Rehabilitation, 87(7), 962–966. https://doi.org/10.1016/j.apmr.2006.02.036

31.

Sainburg

R. L.

(2014). Convergent models of handedness and brain lateralization. Frontiers in Psychology, 5, 1092. https://doi.org/10.3389/fpsyg.2014.01092

32.

Sainburg

R. L.

Schaefer

S. Y.

(2004). Interlimb differences in control of movement extent. Journal of Neurophysiology, 92(3), 1374–1383. https://doi.org/10.1152/jn.00181.2004

33.

Santisteban

Térémetz

Bleton

J.-P.

Baron

J.-C.

Maier

M. A.

Lindberg

P. G.

(2016). Upper limb outcome measures used in stroke rehabilitation studies: A systematic literature review. PLOS ONE, 11(5), e0154792. https://doi.org/10.1371/journal.pone.0154792

34.

Smith

D. B.

Scott

S. H.

Semrau

J. A.

Dukelow

S. P.

(2023). Impairments of the ipsilesional upper-extremity in the first 6-months post-stroke. Journal of NeuroEngineering and Rehabilitation, 20(1), 106. https://doi.org/10.1186/s12984-023-01230-8

35.

Subramaniam

Varghese

Bhatt

(2019). Influence of chronic stroke on functional arm reaching: Quantifying deficits in the ipsilesional upper extremity. Rehabilitation Research and Practice, 2019(1), 5182310. https://doi.org/10.1155/2019/5182310

36.

Thompson-Butel

A. G.

Lin

Shiner

C. T.

McNulty

P. A.

(2015). Comparison of three tools to measure improvements in upper-limb function with poststroke therapy. Neurorehabilitation and Neural Repair, 29(4), 341–348. https://doi.org/10.1177/1545968314547766

37.

Tolle

K. A.

Rahman-Filipiak

A. M.

Hale

A. C.

Kitchen Andren

K. A.

Spencer

R. J.

(2020). Grooved Pegboard Test as a measure of executive functioning. Applied Neuropsychology: Adult, 27(5), 414–420. https://doi.org/10.1080/23279095.2018.1559165

38.

Wetter

Poole

J. L.

Haaland

K. Y.

(2005). Functional implications of ipsilesional motor deficits after unilateral stroke. Archives of Physical Medicine and Rehabilitation, 86(4), 776–781. https://doi.org/10.1016/j.apmr.2004.08.009

39.

Woytowicz

E. J.

Rietschel

J. C.

Goodman

R. N.

Conroy

S. S.

Sorkin

J. D.

Whitall

McCombe Waller

(2017). Determining levels of upper extremity movement impairment by applying a cluster analysis to the Fugl-Meyer assessment of the upper extremity in chronic stroke. Archives of Physical Medicine and Rehabilitation, 98(3), 456–462. https://doi.org/10.1016/j.apmr.2016.06.023