Sage Journals: Discover world-class research

Abstract

Study Design:

Systematic review.

Objectives:

The optoelectronic camera source and data interpolation process serve as the foundation for navigational integrity in robotic-assisted surgical platforms. The current systematic review serves to provide a basis for the numerical disparity observed when comparing the intrinsic accuracy of optoelectronic cameras versus accuracy in the laboratory setting and clinical operative environments.

Methods:

Review of the PubMed and Cochrane Library research databases was performed. The exhaustive literature compilation obtained was then vetted to reduce redundancies and categorized into topics of intrinsic accuracy, registration accuracy, musculoskeletal kinematic platforms, and clinical operative platforms.

Results:

A total of 465 references were vetted and 137 comprise the basis for the current analysis. Regardless of application, the common denominators affecting overall optoelectronic accuracy are intrinsic accuracy, registration accuracy, and application accuracy. Intrinsic accuracy equaled or was less than 0.1 mm translation and 0.1 degrees rotation per fiducial. Controlled laboratory platforms reported 0.1 to 0.5 mm translation and 0.1 to 1.0 degrees rotation per array. Accuracy in robotic-assisted spinal surgery reported 1.5 to 6.0 mm translation and 1.5 to 5.0 degrees rotation when comparing planned to final implant position.

Conclusions:

Navigational integrity and maintenance of fidelity of optoelectronic data is the cornerstone of robotic-assisted spinal surgery. Transitioning from controlled laboratory to clinical operative environments requires an increased number of steps in the optoelectronic kinematic chain and error potential. Diligence in planning, fiducial positioning, system registration and intra-operative workflow have the potential to improve accuracy and decrease disparity between planned and final implant position.

Keywords

optoelectronic accuracy spinal surgery spinal biomechanics imaging navigation robotics

Introduction

The fundamental technological challenge of navigation and robotic-assisted spinal surgery is the virtual world needs to clearly represent the physical, real time world. Among the multiple applications, variables and equipment utilized in navigation and robotic-assisted spinal surgery, the optoelectronic camera source and data interpolation process serves as the foundation for navigational integrity and accuracy, or lack thereof, in the surgical platform. The principles of optoelectronic measurement systems are founded on the basis of devices which have the capability to source, detect and control light and are typically considered a subdivision of photonics. The spectrum of optoelectronic technology platforms are quite diverse, with utilization in sports performance activities such as speed skating and soccer,^1-4 human ergonomics,^5,6 clinical gait and motion analysis,^7-12 musculoskeletal kinematics,^13-20 and clinical operative procedures.^21-33 To this end, the degree of accuracy and errors acceptable across optoelectronic motion measurement platforms differ considerably based on application.¹ For example, fiducial arrays placed on anatomic pelvic landmarks of alpine skiers reported translation accuracy and errors of 8.37 ± 7.1 millimeters (mm).³⁴ Although considered adequate for the evaluation of positional or orientation-related differences in this athletic application, discrepancies of this magnitude would be unacceptable in the clinical operative setting. Technological advancements in the accuracy of optoelectronic marker-based systems over the past 20 years have facilitated the adoption and application of these platforms to the field of robotic-assisted spinal surgery.^12,35,36 An ensuing plethora of journal publications have documented the safety, efficacy and technical accuracy of navigation and robotic systems,^{21-24,30-32,37-45} operative surgical applications,^25,33,46,47 as well as challenges of process workflow, learning curve and training.^{28,30,31,39,40}

Review of these publications, reveals what could be defined as a significant discrepancy when comparing optoelectronic accuracy in the laboratory setting versus clinical operative environment. An approximate 10-fold decrease in technical accuracy of final implant position (≤2 mm) in the clinical operative environment was observed compared to controlled musculoskeletal kinematic studies (≤0.2 mm), despite utilization of near identical optoelectronic camera systems. Hence, the objective of the current systematic review serves to provide a basis for the numerical disparity that exist when comparing intrinsic accuracy of optoelectronic cameras, accuracy observed in the laboratory setting and accuracy in the clinical operative environment. It is postulated that there exists a greater number of linkages in the optoelectronic kinematic chain when analyzing the clinical operative environment to laboratory setting. This increase in data interpolation, coupled with intraoperative challenges, reduces the degree of accuracy compared to that observed in controlled musculoskeletal kinematic laboratory investigations.

Methods

A comprehensive systematic review of the PubMed and Cochrane Library research databases was performed. The time interval was unrestricted, however, the majority of publications comprising the basis of this analysis ranged from 2000–present. Combinations of key search terms were stratified into the following: optoelectronic measurement systems, technical accuracy, experimental error, robotic assisted spinal surgery, spinal kinematics, and navigation. The search was limited to papers in the English language, indexed in peer-reviewed journals accessible through online searches, and all publications included required a bona fide PubMed identification (pmid) or digital object identifier (doi) citation. The exhaustive literature compilation obtained was then pooled in an EndNote file, vetted to reduce redundancies and categorized into topics pertinent to optoelectronic measurement system accuracy with specific reference to intrinsic accuracy, registration accuracy, musculoskeletal kinematic platforms, and clinical operative platforms. The primary tier for inclusion focused on publications which reported quantitative units of measure (microns, millimeters and degrees) for intrinsic camera accuracy and tolerances, accuracy obtained in a controlled laboratory setting and accuracy in the clinical operative setting. Musculoskeletal kinematic measurement studies were included if motions observed were greater than the standard error of measure (SEM). The second tier of inclusion criteria focused on publications highlighting the applications of optoelectronic measurement systems, percent accuracy of screw placement, inherent inaccuracies of surgical instruments, observational commentary leading to accuracy improvement, and methods to mitigate error potential.

Results

Three-Dimensional Cartesian Rigid Body Transformations

The reported optoelectronic measurements of accuracy, errors and methods to quantify these in the laboratory setting or clinical operative environment are based on a fixed 3-dimensional Cartesian coordinate system of rigid body transformation in millimeters (mm) translation and degrees (deg) rotation along 3 orthogonal axes—X, Y and Z.^48-52 This is in accordance to the axial (Y), sagittal (Z) and coronal (X) anatomic planes as defined by Panjabi’s 3-dimensional conceptual framework for spinal kinematics (Figure 1A and B).^53-55 From a nomenclature standpoint, accuracy is defined as a combination of trueness and precision according to the published ISO standard 5725-1.⁵⁶ Trueness refers to the difference between measured value and true position—typically represented by the mean value of repeated measurements. Precision is a measure of repeatability—typically represented by the standard deviation of repeated measurements and refers to random error and noise within the system. In addition to these standardizations, a useful key measure with regard to accuracy (trueness and precision) is the root mean square distance error (RMS) as given by ei being the 3-dimensional distance error of measurement i and N the number of measurements.⁵⁷

Figure 1.

Cartesian coordinate system and conceptual fame for spinal kinematics—schematic representation of a fixed 3-dimensional cartesian coordinate system for calculation of rigid body transformation in millimeters (mm) translation and degrees (deg) rotation along 3 orthogonal axes—X, Y and Z (A). This is in accordance to the axial (Y), sagittal (Z), and coronal (X) anatomic planes as defined by Panjabi’s 3-dimensional conceptual framework for spinal kinematics (B).

e_{R M S} = \sqrt{\frac{1}{N} \sum_{1}^{N} (e_{i} \cdot e_{i})}

Optoelectronic Measurement Systems

Image guided surgery (IGS) is based on the principal of integration and registration of the operative field to pre- or intra-operative data set (e.g. CT or MRI), via amalgamation of an optoelectronic imaging system with robotic platform.⁵⁸ Although not necessarily involved in the execution of operative procedures, optoelectronic measurement systems are considered the gold standard in motion capture accuracy^1,59 and provide 3-dimensional visualization and guidance; improving task execution, targeting accuracy while functioning in a semi-autonomous fashion.⁶⁰ Hence, objective accuracy and error assessments of optoelectronic-robotic interventional platforms is essential. Regardless of optoelectronic camera system, the fundamental triad of common denominators in assessing platform accuracy include: 1) intrinsic accuracy of the source device, 2) registration and tracking accuracy, and 3) application accuracy. Prior to addressing the basis for application accuracy across laboratory versus clinical platforms, the intrinsic and registration accuracy and potential for error propagation are of primary consideration.

Intrinsic (Technical) Accuracy

The initial link in the optoelectronic kinematic chain of data transference resides in the intrinsic accuracy of the camera source. Of the multiple factors affecting downstream optoelectronic accuracy in musculoskeletal kinematic and clinical operative platforms, the intrinsic camera components are most controllable. Mechanical compliance of the system, loose interconnection mechanisms,⁵⁸ variation in camera resolution, calibration, imperfect lenses, number of cameras, spatial orientation, noise, computer vision algorithms and jitter all represent sources of intrinsic error in optoelectronic systems.^36,58,61-63 Topley and Richards³⁶ reported that optoelectronic cameras of higher resolution (Vicon 16MP) and number of cameras (n = 12) significantly improve the 3-dimensional spatial accuracy (0.080 ± 0.092 mm) compared to an equal number of cameras of lower resolution (OptiTrack 1.3MP) (0.259 ± 0.084 mm). Khadem et al⁶⁴ reported the intrinsic optical tracking error secondary to jitter. In the absence of intrinsic jitter, repeated static measurements from a camera source to fiducial would produce identical kinematic signatures along 3 orthogonal axes. Comparison of 5 different optoelectronic camera systems utilized in image based surgical navigation was performed [Image Guided Technologies (IGS), Boulder, CO and Northern Digital Imaging (NDI), Waterloo Canada]. The intrinsic static jitter (mean and standard deviation values) for the IGS Flashpoint systems ranged from 0.028 ± 0.012 mm (Flashpoint 300 mm), 0.051 ± 0.038 mm (FlashPoint 580 mm) to 0.059 ± 0.047 mm (Flashpoint 1 m). The NDI Polaris systems indicated mean values of 0.058 ± 0.037 mm (active LED) and 0.115 ± 0.075 (passive LED). When performing static single marker measurements according to ASTM guidelines⁶⁵ of the NDI Optotrak 3020, a motion analysis system commonly used in musculoskeletal kinematic platforms,^13,17 Maletsky et al⁶⁶ reported the relative accuracy position between 2 rigid bodies at 0.03 mm translation and 0.04 degrees angulation, respectively. Elfring et al⁵⁷ evaluated the static volumetric single marker measurements of 3 commercially available optoelectronic tracking systems utilized in robotic assisted platforms—NDI Polaris P4, NDI Polaris Spectra (active and passive mode) and Stryker Navigation System II (Stryker Inc., Kalamazoo, MI). The Stryker Navigation System II camera and the Polaris Spectra (active mode), exhibited trueness values of 0.058 ± 0.033 mm and 0.089 ± 0.061 mm, respectively. The Polaris Spectra (passive mode) exhibited values of 0.170 ± 0.090 mm, and the Polaris P4 demonstrated the highest static measurement error of 0.272 ± 0.394 mm. As a baseline statement of comparison, the optoelectronic systems utilized in clinical operative or controlled laboratory platforms report only marginal differences in accuracy. Further, the contribution of intrinsic errors is of miniscule value in comparison to error(s) propagation secondary to registration, targeting tracking and application in controlled experimental and clinical operative platforms.

Registration Accuracy and Target Tracking

A second key step in the optoelectronic kinematic chain and highest probable link(s) of error propagation is the registration process. This intra-operative process integrates correlation and mapping algorithms to register the physical patient to the virtual patient via the navigation system, optoelectronic source, fiducial arrays in the operative field, and patient CT images. Accurate, close-to-ideal reference reproducibility of the dataset improves trueness and precision of subsequent intra-operative tracking. Multiple factors affect registration accuracy and target tracking, including optoelectronic camera source, passive versus active arrays, occlusions, distance between fiducial arrays and camera source, and anatomic locations of the coordinate reference fiducials.^1,57,67-74

As reported by States and Pappas⁷⁵ and Simoes et al,⁶⁸ the NDI Certus 3 camera optoelectronic system demonstrated a significantly higher degree of registration target accuracy (0.1 mm translation and 0.13 degrees rotation⁷⁵ compared to the NDI Spectra 2 camera system when located 1.5 meters from camera source.⁶⁸ The surgical navigation principle of triangulation, necessary to quantify 3-dimensional fiducial array position, is void if 1 of 2 cameras is occluded.^61,76,77 Marinetto et al⁷⁸ reported tracking errors secondary to camera and fiducial array occlusions when utilizing an 8-camera configuration. Occlusions of a single camera resulted in tracking errors from 0.2 mm to 0.5 mm, while occlusion of 5 cameras resulted in errors from 0.6 mm to 1.6 mm. Optical trackers utilizing multi-camera systems may lead to data redundancy but enables fidelity in data transference, despite fiducial occlusions within the application platform (Figure 2A-C).

Figure 2.

Optoelectronic camera sources—comparison of 3 optoelectronic camera systems utilized in motion analysis. The NDI spectra 2 camera system (A), NDI certus optotrak 3 camera system (B), and viconvicon MX13 multiple camera system (C) (Vicon motion systems Ltd., Oxford, UK). The spectra NDI is commonly used in the clinical operative environment and latter 2 systems for musculoskeletal kinematics and biomechanics. The Vicon camera image and testing setup was generously provided by Prof. Dr. Hans Joachim Wilke, PhD, Institute of Orthopaedic Research and Biomechanics, University of Ulm.

The difference between active versus reflective passive markers systems also effect registration accuracy and target tracking. In the case of passive markers, the optoelectronic source floods the operative field and light is reflected back to the sensors via infrared-reflecting spheres. Active markers contain and emit infrared-emitting diodes (IREDs) and provide a more robust signal with increased accuracy compared to passive markers.^57,72,79 Furthermore, the distance between optoelectronic camera source and operative field fiducial arrays effect accuracy.⁶⁴ Increasing the camera distance from 6 to 8 feet nearly triples the intrinsic registration error along the Z axis (maximum = 0.250 mm) for the Polaris passive fiducial array system. Hence, closer approximation of the optoelectronic camera source to the operative fiducial arrays (≤6 feet or approximately 1800 millimeters) minimizes jitter and improves precision.⁶⁴ For comparison, when performing static single marker measurements of the NDI Optotrak 3020 system, the precision in rotation degrades significantly when positioning the camera ≥ 2.5meters from fiducial arrays^66,75 (Figure 3A and B).

Figure 3.

Active versus passive marker arrays—comparison of active marker arrays used in the laboratory platform as shown attached to the vertebral elements. Note the active markers contain and emit infrared-emitting diodes (IREDs) via the attached wiring configurations (A). In the case of passive fiducial marker arrays (B), the optoelectronic source floods the operative field and light is reflected back to the camera source via infrared-reflecting spheres.

In addition to distance between camera source and fiducial arrays, Citak et al^70,80 performed an investigation to determine the maximum acceptable distance between fixed reference arrays and dynamic “mobile” arrays within the operative field. The results demonstrated a mean registration error of 0.04 mm (0.04-0.05 mm) up to a distance of 200 mm from the patient reference array. When mobile fiducial arrays (e.g., end effector) exceeded 200 mm from the reference point, the registration error increased to 0.25 mm (0.24-0.26) (P < .0001). The initial registration and intra-operative working accuracy is significantly reduced with localization of arrays greater than 200 mm from the patient reference point.^70,80 Moreover, the magnitude of errors secondary to array obstructions differ based on active versus passive markers and distance between arrays.³⁵ When comparing 2 active Vicon fiducial arrays versus distance, targeting errors increased from 0.47 mm to 1.2 mm, with corresponding stepwise decreases from 5 cm to 0 cm between arrays. In similarity but of greater magnitude, targeting errors for passive arrays (Qualisys AB, Goteborg Sweden) increased from 0.54 mm to 2.99 mm with decreases from 5 cm to 0 cm between arrays. Hence, as visualized by the camera source, increasing distances between fiducial arrays in the operative clinical setting reduces experimental targeting errors (Figure 4A-D).

Figure 4.

Fiducial array occlusions—Schematic illustration comparing fiducial array configuration and occlusions secondary to optimal optoelectronic camera angle, steep camera angle, occluded sphere and interference between 2 fiducial arrays (A-D).

The effect of transitioning from static to dynamic array localization directly influences accuracy and targeting errors.^44,45,69,81 A comparison of static and dynamic motion by Chassat and Lavallee⁶⁹ demonstrated a significant increase in array translation error when moving across a spectrum of conditions located 2 meters from the camera source. Using the best-in-class NDI Optotrak 3020 system, the static position error measured 0.28 mm, static position hand held 0.400 mm, and dynamic measure in translation at constant speed 1.29 mm. The magnitude in angular errors was less than translation but still increased significantly through dynamic localization of instrument arrays.⁶⁹ These findings were corroborated by Stancic et al⁸¹ with static registration indicated errors of 0.11 mm while dynamic motion of the arrays resulted in errors ranging from 0.250 mm to 1.10 mm with increasing velocities.⁸¹ Hence, higher displacement rates of the operative instrument arrays directly affect the camera’s tracking accuracy.

Two additional clinical factors influencing registration accuracy and targeting errors include the tool to tip distance and accuracy of pre- or intra-operative radiographic CT/MRI data.^72,82-84 Static accuracy for a given marker array is on the order of 0.2 mm, however, errors increase significantly when extrapolated to the instrument tip, Wiles et al⁷² reported that increasing the distance between fiducial array to instrument tip from 0 to 100 mm decreases static accuracy from 0.4 mm to 0.85 mm, respectively.⁷² The array to tool tip distance on instruments utilized in robotic assisted platforms far exceed 100 mm (4 inches). In consideration of the linearity of Wiles et al accuracy degradation with distance calculations,⁷² it is postulated that experimental targeting errors of longer instruments are of increased magnitude compared the reported 0.85 mm. Computed tomographic data input affects target accuracy and can be improved with resolutions of 1.0 mm and 2.0 mm versus 3.0 mm slice thickness.^82,83

In summary, propagation of computational measurement errors in the optoelectronic kinematic chain has a compounding effect for the following transitions: 1) Measurement of the intrinsic image plane error (IPE) secondary to errors within the optoelectronic system, 2) transitioning from image error to fiducial location error (FLE), and 3) transition from fiducial location error to tracking target error (TRE). The mathematical expressions for these computational transformations are beyond the scope of the current publication but well documented by Fitzpatrick et al^71,73 and Sielhorst et al.⁶¹ The margins of error secondary to intrinsic and registration accuracy in optoelectronics are more manageable compared to unpredictable factors related to application in the laboratory versus dynamic clinical intraoperative environments.

Application Accuracy—Musculoskeletal Kinematics Laboratory Platform

Over the last 30 years, a plethora of spine publications have documented the biomechanical properties of the occipitocervical through lumbopelvic spine under controlled laboratory conditions utilizing Panjabi’s 3-dimensional conceptual framework for testing.^17,53-55,85 In contrast to the challenges of the clinical operative environment, motion analysis of spinal implant and anatomic vertebral structures(s) in the controlled laboratory setting is performed utilizing a 6 degree of freedom musculoskeletal simulator interfaced with an optoelectronic measurement system. The fundamental principles pertinent to maximizing optoelectronic accuracy include mounting the specimen to a rigid testing platform, affixing active or passive fiducial arrays directly to implants or anatomic structures using screw-bolt fixation, and creating rigid body configurations parallel to the camera source. Multidirectional flexibility testing is typically performed along 3 predominant loading axes—flexion-extension, lateral bending, and axial rotation under controlled displacement rates of 1 to 3 degrees per second for multiple cycles. To this end, a series of laboratory investigations using the NDI Certus and Vicon MX13 camera systems (Vicon Motion Systems Ltd., Oxford, UK) reported the peak limits of optoelectronic accuracy when evaluating kinematics of the osteoligamentous spine^13-19,86-92 (Figure 5A and B).

Figure 5.

Application accuracy—musculoskeletal kinematics laboratory platform in contrast to the challenges of the clinical operative environment, motion analysis of spinal implant and anatomic vertebral structures(s) in the controlled laboratory setting is performed utilizing a 6 degree of freedom musculoskeletal simulator interfaced with an optoelectronic measurement system (A). The fundamental principles pertinent to maximizing optoelectronic accuracy include mounting the specimen to a rigid testing platform, affixing active or passive fiducial arrays directly to implants or anatomic structures using screw-bolt fixation, and creating rigid body configurations parallel to the camera source (B).

Cunningham et al¹⁴ compared occipital plate versus intracranial anchors for reconstruction of the occipitocervical (O-C) junction. The reported differences (degrees) in axial rotation at the O-C junction based on optoelectronic measurements were 4.13 ± 2.05 (intact), 0.22 ± 0.13 (plate) and 0.30 ± 0.21 (anchor). Rotation of the plate and anchor, with respect to the occiput, in flexion-extension ranged from 0.06 ± 0.05 to 0.10 ± 0.08, respectively. Although not of clinical significance, the study quantified differences on the order of 0.1 degree between 2 methods of occipitocervical fixation. Bowden et al⁸⁶ described the variability in “quality of motion” of L4 relative to the L5 vertebral elements, following various methods of vertical preload. The optoelectronic data quantified differences in L4 anteroposterior translations of 0.6 mm and rotations of 0.6 degrees between different testing conditions. Ilharreborde et al⁸⁷ reported dynamic kinematic evaluation of the multi-segmental lumbar spine under dynamic loading conditions indicated peak accuracy to within 1.10 ± 0.18 degrees (L2-L3) and corresponding translations of 0.48 ± 0.06 mm rotation (L3-L4) with application of 7.5 Nm pure moment load. In a complex kinematic study utilizing a Vicon optoelectronic system, La Barbera et al^18,19 investigated lumbar interbody cages with Ponte osteotomy versus pedicle subtraction osteotomy for severe sagittal imbalance, the peak accuracy of neutral zone measurements (degrees) across the intact L3-L5 segments demonstrated values of 0.7 (Range 0.3-1.9) in flexion-extension, 1.0 (Range 0.1-3.8) in lateral bending and 0.2 (Range 0.1-0.9) in axial rotation in flexion-extension.

From a kinematic standpoint, the sacroiliac junction (SIJ) presents formidable challenge to definitive and accurate measurement using optoelectronics—reaching the 0.1 degree accuracy error measurement limitations of most optoelectronic systems.^18,19 Jeong et al⁸⁸ was able to differentiate the range of motion (degrees) of the SIJ in lateral bending when comparing the intact condition (1.5 ± 1.5), unilateral fusion (1.4 ± 1.6) and bilateral fusion (1.1 ± 1.0). Despite quantification of SIJ motion within a range of 0.5 degrees of accuracy, the comparisons were not significant. Osterhoff et al⁸⁹ quantified the effects of cement augmentation on sacroiliac screw position and fracture site motion using NDI optoelectronics. Screw tip positioning within the sacrum was quantified to an accuracy level of 0.7 mm (Range 0.5-1.3), with a corresponding vertical (Y axis) SIJ range of motion of 1.2 mm (Range 0.6-1.9) under cyclic compressive loads. In a comprehensive S2 alar iliac screw instrumentation study using 21 lumbopelvic specimens, Cunningham and co-workers¹³ reported SIJ motion (degrees) to accuracies of 1.78 ± 0.96 (flexion-extension), 0.52 ± 0.34 (lateral bending) and 0.48 ± 0.32 degrees (axial rotation). In a similar study by Dall et al¹⁵ evaluating SIJ fusion using lateral sacroiliac screws, optoelectronic tracking quantified the intact SIJ motion (degrees) under flexion-extension, lateral bending and axial rotation at 1.80 ± 1.62, 0.44 ± 0.35 and 1.13 ± 0.82, respectively. Additional studies on intact SIJ kinematics reported translations along the 3-dimensional coordinate axes on the order of 0.1 mm to 1.04 mm with corresponding rotations of 0.11 to 1.14 degrees.^16,91,92

Laboratory workflow methods and conditions for experimental musculoskeletal kinematic studies are streamlined and optimized for maximizing optoelectronic accuracy. Factors of specimen stabilization, alignment, camera resolution, proximity to fiducials, planar visualization of the active arrays, and controlled motion application account for the high degree of accuracy reported in these studies. Moreover, in difference to the operative clinical environments, non-destructive testing procedures can be repeated multiple times on the same specimen to improve data accuracy. The collective effect of testing methodology and limited experimental coordinate transformations between data input/output reduces error propagation and maximizes optoelectronic accuracy (Figure 6).

Figure 6.

Laboratory platform for optoelectronic data transference process—Schematic illustration demonstrating the laboratory workflow and process for data transference utilizing optoelectronic tracking. The camera source visualizes the active fiducial arrays affixed to the vertebral elements and transfers the data directly to the user interface for computational analysis. The collective effect of testing methodology and limited experimental coordinate transformations between data input/output reduces error propagation and maximizes optoelectronic accuracy.

Application Accuracy—Clinical Operative Platform

Transitioning from controlled laboratory conditions to the dynamic variability of a clinical operative environment presents a different set of application challenges for maintaining peak optoelectronic accuracy. Unique to robotic assisted spinal surgery and in difference to the laboratory setting, the intra-operative process requires considerably more steps in the transference of optoelectronic kinematic data. This complex process flow integrates correlation and mapping algorithms to register the physical patient to the virtual patient via the navigation system, optoelectronic source, surveillance markers, patient reference markers, end effector instruments in the operative field, and patient CT images. Accurate, close-to-ideal reference reproducibility and maintenance of this dataset is the primary intra-operative objective and challenge (Figure 7).

Figure 7.

Application accuracy—clinical operative platform. Intra-operative images highlight the transition from controlled laboratory conditions to the dynamic variability of the clinical operative environment (A). A different set of application challenges are required for maintaining peak optoelectronic accuracy related to registering the physical patient to the virtual patient via the navigation system, optoelectronic source, surveillance markers, patient reference markers, end effector instruments in the operative field and workflow. The intra-operative images were generously provided by the operative surgeons, (A) Dr. Vladimir Sinkov, MD and (B) Dr. Bhavuk Garg, MD.

Despite utilization of near identical optoelectronic sources and fiducial arrays, a consistent disparity exists when comparing the reported technical accuracies in the laboratory setting versus clinical operative environment. An approximate 10-fold decrease in accuracy was observed when comparing the final implant position (≤2 mm) in the clinical operative environment to musculoskeletal kinematic studies (≤0.2 mm). An extensive number of peer reviewed journal publications have documented the use, efficacy safety and technical accuracy achieved with robotic-assisted spinal surgery.^{23,28,30,37,39-41,44-46,93-123} The focus in reviewing these publications was to highlight the technical accuracy observations and determine a basis for discrepancy between planned versus actual final implant position based on postoperative CT images. In case studies where quantitative measurements were not reported, the Gertzbein and Robbins¹²⁴ score (GRS) was adopted to calculate pedicle screw implant position. According to the GRS classification, screws centered within the pedicle are considered grade A; < 2 mm from center is a grade B; a breach from 2 mm to 4 mm is grade C; a breach from 4 mm to 6 mm is grade D; and > 6 mm is grade E. Grades of A and B (< 2 mm pedicle breach) are considered clinically acceptable, and all other grades indicate malposition.

Helm et al⁹⁴ performed a comprehensive literature review on the technical accuracy of 12 622 pedicle screws implanted using a variety of image guided surgery navigation systems. As reported, 11 830 were positioned perfectly according to pre-operative plan (A), 395 screws within less than 2 mm of plan (B), 92 breached between 2 mm to 4 mm off center (C), and 55 were within 4 mm to 6 mm of the pre-operative plan. The balance of 250 screws remained ungraded due to radiographic issues. Zhang et al⁴⁵ reported one of the largest compilations assessing screw accuracy in robotic assisted spinal surgery. A total of 23 studies including prospective, prospective randomized control trials and retrospective reviews comprised the basis of this publication with a total of 5,013 pedicle screw positions evaluated. The accuracy according to GRS grades of A and B (less than 2 mm cortical breach) was 4,781 screws (95.38%), while the balance of 232 screws (4.61%) ranged from 2 mm to 6 mm when comparing the planned to actual final positions. Solomiichuk et al⁴¹ reported on 192 screws implanted in 35 patients. Trajectories were Grade A or B (less than 2 mm) in 162 (84.4%) of screws. The malposition rate of 2 mm to 6 mm was present in 30 of 192 screws (15.6%) with 23 of these occurring in the thoracic spine, where pedicle widths are significantly less than the lumbar region.^103 125 Devito et al⁹³ reported on the technical accuracy placement of 646 pedicle screws inserted in 139 patients using postoperative CT scans. 577 were centered in the pedicle, 58 were less than 2 mm off center, 9 breached 2 mm to 4 mm off center and 2 screws deviated greater than 4 mm from the pedicle wall. Schatlo et al³⁹ reported on the technical accuracy of 244 lumbar pedicle screws. 204 screws (83.6%) were graded as a perfect trajectory (A) compared to the pre-operative plan, 19 (7.8%) were less than 2 mm (B), 9 (3.7%) breached 2 mm to 4 mm off center (C), 4 (1.6%) breached from 4 mm to 6 mm (D), 2 (0.8%) were greater than 6 mm off center and 6 (2.5%) screws required revision.

A comprehensive study from Keric et al¹⁰⁷ reviewed the technical accuracy of 1857 screw positions based on postoperative CT scans. Of the 1857 screws, 1799 (96.9%) were graded as acceptable or good position, 38 screws (2%) exhibited deviations from 3 mm to 6 mm and 20 screws (1.1%) were greater than 6 mm from the planned trajectory. The 58 malpositioned screws (3 mm to 6 mm deviations) were located primarily in the upper and lower thoracic regions versus the lumbar spine. These deviations were considered secondary to smaller pedicle morphometry¹²⁶ and instrument skiving.¹⁰⁷

In consideration of the range for “acceptable” deviations (best case = <2 mm) between planned and actual positions in the clinical setting, attention must be directed toward the anatomical pedicle morphology from the cervical to lumbar regions. For typical cervical vertebrae (C3 to C6) the mean pedicle width was reported 4.9 ± 0.9 mm.^127,128 In the thoracic region, pedicle width progressively decreases from T1 to T5 (mean 3.65 ± 0.40 mm) and increases from T6 though the T12 levels (mean 7.89 ± 0.70 mm).^103,125 The lumbar vertebrae allow greater cross sectional areas for “ideal” pedicle screw placement, however, the proximity of adjacent exiting nerve roots ranges from 2.9 mm to 6.2 mm proximally and 0.8 mm to 2.8 mm distally. Moreover, the distance between the medial borders of pedicle to thecal sac ranges from 0.9 mm to 2.1 mm.¹²⁹

Basis for Disparity in Optoelectronic Accuracy

A key consideration pertaining to optoelectronic accuracy in the clinical environment compared to the laboratory setting is the dynamic nature of the operating room. The basis for disparity in accuracy when equating the laboratory versus clinical operative platforms is a result of the combined, cumulative errors secondary to the intraoperative workflow process, variability in anatomic morphology, and spinal flexibility. Of fundamental importance, and the crux of the matter, related to error propagation in navigation and robotic-assisted spinal surgery is the assumption that the workflow platform and patient’s spine is rigid, and as such, motion of any type is perceived as a rigid body transformation. Optoelectronic error reduction in the clinical flow requires stabilization of the camera source, rigid fixation of surveillance arrays in the iliac crest, stable attachment of patient reference and registration arrays to anatomic landmarks, and end effector instruments arrays which are inflexible. The end effector is the last link where the robotic enters the workspace and small rotations or translations in the array references can lead to large errors in instrument position. Although accurate, close-to-ideal reference reproducibility of these steps will reduce errors, the reality is that fixed arrays do move—leading to increased relative motions between arrays and subsequent error propagation and disparity between the physical, real time world and virtual world. Moreover, spatial errors can be further magnified due to geometrical distortion of preoperative images, and tracking error of the surgical instruments.¹³⁰ To register the physical patient to the virtual patient, Grunert et al⁶⁷ proposed a series of transformation matrices, including fiducial-based paired-point transformation, surface contour matching, and hybrid transformation. The hybrid transformation process is most applicable to robotic assisted spinal surgery as it includes the methods of surface-based and pair-point based methods with implanted fiducials. As such, tracing at least 3 anatomic landmarks with navigational confirmation serves to reduce error potential.

Several publications on optimizing clinical workflow process have been reported^{31,38,42,76 130-133} Lieberman et al³¹ provides an excellent description of the step-by-step workflow process in robotic assisted spinal surgery. The report provides a concise methodological approach to operative workflow, while at the same time providing a collective basis for potential error(s) propagation in the clinical setting. The sequential description of process flow/error potentials includes pre-operative and intra-operative registration, dislodgement of reference arrays, damaged or bent navigation tools^134-136 and arrays occlusions (e.g., distance and blood), skiving or tool deflection secondary to sloped anatomic topology or muscle retraction, and untracked patient movement during the spinal destabilizing procedure. In addition to unintended motion or bending of fiducial arrays, the inherent differences in anatomic topology, bone mineral density and flexibility of the patient’s spine, both before and following destabilization and reconstruction procedures, cannot be overemphasized. The challenge is the spine is often flexible—the drill and robotic arm may be properly located, but highly mobile, multi-segmental spinal reconstructions with minimal deflection force leads to unintended rotation or translation of the operative vertebral elements, skiving or tool deflection, and effects precision rate during screw insertion.^{31,38 132 133 135 137} The basis for decreased technical accuracy in the clinical operative platforms is a result of combined, cumulative errors secondary to the intraoperative workflow process, number of kinematic linkages and variability in patient spinal morphology and flexibility (Figure 8).

Figure 8.

Clinical platform for optoelectronic data transference process. Schematic illustration demonstrating the operative clinical workflow and process for data transference utilizing optoelectronic tracking. Unique to robotic assisted spinal surgery and in difference to the laboratory setting, the intra-operative process requires considerably more steps in the transference of optoelectronic kinematic data. This complex workflow process integrates correlation and mapping algorithms to register the physical patient to the virtual patient via the navigation system, optoelectronic source, surveillance markers, patient reference markers, end effector instruments in the operative field, and patient CT images. Accurate, close-to-ideal reference reproducibility and maintenance of this dataset is the primary intra-operative objective and challenge.

Discussion

In reviewing the intrinsic technical accuracy and registration accuracy, there exists a substantial burden of proof that the potential performance in optoelectronics is nearly identical between the 2 platforms—laboratory versus clinical operative—under static conditions. The downstream difference in optoelectronic technical accuracy and disparity between the 2 platforms is secondary to the dynamic factors unique to each. The laboratory workflow methods and array registration for experimental musculoskeletal kinematic studies are rigid, highly controlled, with limited experimental coordinate transformations between data input/output—reducing error propagation and maximizing optoelectronic accuracy. Unique to robotic-assisted spinal surgery and in difference to the laboratory setting, the dynamic intra-operative process necessitates considerably more steps in the transference of optoelectronic kinematic data. The complex data flow process integrates correlation and mapping algorithms to register the physical patient to the virtual patient via the navigation system, optoelectronic source, surveillance markers, patient reference markers, end effector instruments, and patient CT images. Essentially, this is a comparison of technical accuracy between a rigid, highly controlled setting and a variable environment with multiple data input factors. The collective effect results in an increased potential for error propagation from experimental coordinate transformations, data processing and optoelectronic kinematic linkages in the clinical setting.

Although workflow and patient related factors provide a basis for decreased accuracy in the clinical setting, with differences between planned versus actual final implant position ranging from 1.5 mm to 3 mm, it could be argued with some degree of confidence that these technical inaccuracies are inconsequential and of no clinical significance in anatomic zones permitting such deviation (e.g., L5 pedicle). However, in cases of cervical or thoracic operative procedures with pedicular dimensions less than 3.5 mm diameter and neural structures within 0.8 mm proximity to pedicular cortices, technical errors of 2 mm to 3 mm are significant, and navigational integrity and reliability of the data transformation process are of paramount importance. With this level of technical inaccuracy, there is basis for contraindication in the use of navigation and robotic-assisted spinal surgery depending on the indications presented and extent of confounding factors that may decrease registration accuracy and subsequent technical accuracy of implant placement.

The fundamental technological challenge of navigation and robotic-assisted spinal surgery is the virtual world needs to clearly represent the physical, real time world. Navigational integrity and maintenance of fidelity in the transference of optoelectronic data is the cornerstone of robotic-assisted spinal surgery. Transitioning from the controlled laboratory setting to clinical operative environment requires an increased number of steps in the optoelectronic kinematic chain and potential for error propagation in experimental coordinate transformations. Moreover, intra-operative challenges of array location, system registration, spinal flexibility, anatomic topography and workflow affect navigational integrity and provide a basis for the disparity of optoelectronic accuracy in the clinical environment compared to the controlled laboratory setting. A continuum of decreased accuracy is demonstrated when comparing the optoelectronic camera source itself to application in musculoskeletal platforms, and finally, clinical operative environment. Diligence in the areas of pre-operative planning, source camera and fiducial positioning, system registration and intra-operative process workflow have the potential to improve accuracy and decrease disparity between planned and final implant position.

Footnotes

Acknowledgments

The authors would like to credit Mir Hussain, Nobert Johnson (Globus Medical Corporation), and Dr. Vladimir Sinkov, MD, for their expertise and objective comments in the preparation of this manuscript.

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Bryan W. Cunningham, PhD

References

van der Kruk

Reijne

. Accuracy of human motion capture systems for sport applications; state-of-the-art review. Eur J Sport Sci. 2018;18(6):806–819. doi:10.1080/17461391.2018.1463397

Linke

Link

Lames

. Validation of electronic performance and tracking systems EPTS under field conditions. PLoS One. 2018;13(7):e0199519. doi:10.1371/journal.pone.0199519

Linke

Link

Lames

. Football-specific validity of TRACAB’s optical video tracking systems. PLoS One. 2020;15(3):e0230179. doi:10.1371/journal.pone.0230179

Elliott

Alderson

. Laboratory versus field testing in cricket bowling: a review of current and past practice in modelling techniques. Sports Biomech. 2007;6(1):99–108. doi:10.1080/14763140601058623

Lindbeck

Kjellberg

. Gender differences in lifting technique. Ergonomics. 2001;44(2):202–214. doi:10.1080/00140130120142

Nimbarte

Sun

Jaridi

Hsiao

. Biomechanical loading of the shoulder complex and lumbosacral joints during dynamic cart pushing task. Appl Ergon. 2013;44(5):841–849. doi:10.1016/j.apergo.2013.02.008

Negrini

Piovanelli

Amici

, et al. Trunk motion analysis: a systematic review from a clinical and methodological perspective. Eur J Phys Rehabil Med. 2016;52(4):583–592.

Bravi

Gallotta

Morrone

, et al. Concurrent validity and inter trial reliability of a single inertial measurement unit for spatial-temporal gait parameter analysis in patients with recent total hip or total knee arthroplasty. Gait Posture. 2020;76:175–181. doi:10.1016/j.gaitpost.2019.12.014

Serrao

Casali

Ranavolo

, et al. Use of dynamic movement orthoses to improve gait stability and trunk control in ataxic patients. Eur J Phys Rehabil Med. 2017;53(5):735–743. doi:10.23736/S1973-9087.17.04480-X

10.

Baker

. Gait analysis methods in rehabilitation. J Neuroeng Rehabil. 2006;3(1):4. doi:10.1186/1743-0003-3-4

11.

Andújar

Ribeiro

Fernández-Quintanilla

Dorado

. Accuracy and feasibility of optoelectronic sensors for weed mapping in wide row crops. Sensors (Basel). 2011;11(3):2304–2318. doi:10.3390/s110302304

12.

Colyer

Evans

Cosker

Salo

AIT

. A review of the evolution of vision-based motion analysis and the integration of advanced computer vision methods towards developing a markerless system. Sports Med Open. 2018;4(1):24. doi:10.1186/s40798-018-0139-y

13.

Cunningham

Sponseller

Murgatroyd

Kikkawa

Tortolani

. A comprehensive biomechanical analysis of sacral alar iliac fixation: an in vitro human cadaveric model. J Neurosurg Spine. 2019;30(3):367–375. doi:10.3171/2018.8.SPINE18328

14.

Cunningham

Mueller

Mullinix

Sun

Sandhu

. Biomechanical analysis of occipitocervical stabilization techniques: emphasis on integrity of osseous structures at the occipital implantation sites. J Neurosurg. 2020;33(2):138–147. doi:10.3171/2020.1.spine191331

15.

Dall

Eden

Cho

, et al. Biomechanical analysis of motion following sacroiliac joint fusion using lateral sacroiliac screws with or without lumbosacral instrumented fusion. Clin Biomech (Bristol, Avon). 2019;68:182–189. doi:10.1016/j.clinbiomech.2019.05.025

16.

Mushlin

Brooks

Olexa

, et al. A biomechanical investigation of the sacroiliac joint in the setting of lumbosacral fusion: impact of pelvic fixation versus sacroiliac joint fixation. J Neurosurg Spine. 2019;31(4):1–6. doi:10.3171/2019.3.SPINE181127

17.

Oxland

. Fundamental biomechanics of the spine—what we have learned in the past 25 years and future directions. J Biomech. 2016;49(6):817–832. doi:10.1016/j.jbiomech.2015.10.035

18.

La Barbera

Wilke

Liebsch

, et al. Biomechanical in vitro comparison between anterior column realignment and pedicle subtraction osteotomy for severe sagittal imbalance correction. Eur Spine J. 2020;29(1):36–44. doi:10.1007/s00586-019-06087-x

19.

La Barbera

Wilke

Ruspi

, et al. Load-sharing biomechanics of lumbar fixation and fusion with pedicle subtraction osteotomy. Sci Rep. 2021;11(1):3595. doi:10.1038/s41598-021-83251-8

20.

Camino Willhuber

Zderic

Gras

, et al. Analysis of sacro-iliac joint screw fixation: does quality of reduction and screw orientation influence joint stability? A biomechanical study. Int Orthop. 2016;40(7):1537–1543. doi:10.1007/s00264-015-3007-0

21.

Wang

Feng

, et al. Percutaneous sacroiliac screw placement: a prospective randomized comparison of robot-assisted navigation procedures with a conventional technique. Chin Med J (Engl). 2017;130(21):2527–2534. doi:10.4103/0366-6999.217080

22.

Wang

Shang

Wang

. A review of computer-assisted orthopaedic surgery systems. Int J Med Robot. 2020;16(5):1–28. doi:10.1002/rcs.2118

23.

D’Souza

Gendreau

Feng

Kim

Veeravagu

. Robotic-assisted spine surgery: history, efficacy, cost, and future trends. Robot Surg. 2019;6:9–23. doi:10.2147/RSRR.S190720

24.

Staub

Sadrameli

. The use of robotics in minimally invasive spine surgery. J Spine Surg. 2019;5(Suppl 1):S31–S40. doi:10.21037/jss.2019.04.16

25.

Kim

Jung

Chang

Lee

Kang

Yeom

. A prospective, randomized, controlled trial of robot-assisted vs freehand pedicle screw fixation in spine surgery. Int J Med Robot. 2017;13(3):e1779. doi:10.1002/rcs.1779

26.

Han

Tian

Liu

, et al. Safety and accuracy of robot-assisted versus fluoroscopy-assisted pedicle screw insertion in thoracolumbar spinal surgery: a prospective randomized controlled trial. J Neurosurg Spine. 2019:1–8. doi:10.3171/2018.10.SPINE18487

27.

Ahern

Gibbons

Schroeder

Vaccaro

Butler

. Image-guidance, robotics, and the future of spine surgery. Clin Spine Surg. 2020;33(5):179–184. doi:10.1097/BSD.0000000000000809

28.

Lieber

Kirchner

Kerbel

Khalsa

. Robotic-assisted pedicle screw placement fails to reduce overall postoperative complications in fusion surgery. Spine J. 2019;19(2):212–217. doi:10.1016/j.spinee.2018.07.004

29.

Sayari

Pardo

Basques

Colman

. Review of robotic-assisted surgery: what the future looks like through a spine oncology lens. Ann Transl Med. 2019;7(10):224. doi:10.21037/atm.2019.04.69

30.

Lieberman

Togawa

Kayanja

, et al. Bone-mounted miniature robotic guidance for pedicle screw and translaminar facet screw placement: part I—technical development and a test case result. Neurosurgery. 2006;59(3):641–650. Discussion 641-50. doi:10.1227/01.neu.0000229055.00829.5b

31.

Lieberman

Kisinde

Hesselbacher

. Robotic-assisted pedicle screw placement during spine surgery. JBJS Essent Surg Tech. 2020;10(2):e0020. doi:10.2106/JBJS.ST.19.00020

32.

Shoham

Lieberman

Benzel

, et al. Robotic assisted spinal surgery—from concept to clinical practice. Comput Aided Surg. 2007;12(2):105–115. doi:10.3109/10929080701243981

33.

Ghasem

Sharma

Greif

Alam

Maaieh

. The arrival of robotics in spine surgery: a review of the literature. Spine (Phila Pa 1976). 2018;43(23):1670–1677. doi:10.1097/BRS.0000000000002695

34.

Spörri

Schiefermüller

Müller

. Collecting kinematic data on a ski track with optoelectronic stereophotogrammetry: a methodological study assessing the feasibility of bringing the biomechanics lab to the field. PLoS One. 2016;11(8):e0161757. doi:10.1371/journal.pone.0161757

35.

Richards

JG.

The measurement of human motion: a comparison of commercially available systems. Hum Mov Sci. 1999;18(5):589–602. doi:10.1016/s0167-9457(99)00023-8

36.

Topley

Richards

. A comparison of currently available optoelectronic motion capture systems. J Biomech. 2020;106:109820. doi:10.1016/j.jbiomech.2020.109820

37.

Kantelhardt

Martinez

Baerwinkel

Burger

Giese

Rohde

. Perioperative course and accuracy of screw positioning in conventional, open robotic-guided and percutaneous robotic-guided, pedicle screw placement. Eur Spine J. 2011;20(6):860–868. doi:10.1007/s00586-011-1729-2

38.

Lonjon

Chan-Seng

Costalat

Bonnafoux

Vassal

Boetto

. Robot-assisted spine surgery: feasibility study through a prospective case-matched analysis. Eur Spine J. 2016;25(3):947–955. doi:10.1007/s00586-015-3758-8

39.

Schatlo

Molliqaj

Cuvinciuc

Kotowski

Schaller

Tessitore

. Safety and accuracy of robot-assisted versus fluoroscopy-guided pedicle screw insertion for degenerative diseases of the lumbar spine: a matched cohort comparison. J Neurosurg Spine. 2014;20(6):636–643. doi:10.3171/2014.3.SPINE13714

40.

Schatlo

Martinez

Alaid

, et al. Unskilled unawareness and the learning curve in robotic spine surgery. Acta Neurochir (Wien). 2015;157(10):1819–1823. doi:10.1007/s00701-015-2535-0

41.

Solomiichuk

Fleischhammer

Molliqaj

, et al. Robotic versus fluoroscopy-guided pedicle screw insertion for metastatic spinal disease: a matched-cohort comparison. Neurosurg Focus. 2017;42(5):E13. doi:10.3171/2017.3.FOCUS1710

42.

Roser

Tatagiba

Maier

. Spinal robotics: current applications and future perspectives. Neurosurgery. 2013;72(Suppl 1):12–18. doi:10.1227/NEU.0b013e318270d02c

43.

Vadalà

De Salvatore

Ambrosio

Russo

Papalia

Denaro

. Robotic spine surgery and augmented reality systems: a state of the art. Neurospine. 2020;17(1):88–100. doi:10.14245/ns.2040060.030

44.

Zhang

, et al. Multiple instruments motion trajectory tracking in optical surgical navigation. Opt Express. 2019;27(11):15827–15845. doi:10.1364/OE.27.015827

45.

Zhang

Han

, et al. Robotic navigation during spine surgery. Expert Rev Med Devices. 2020;17(1):27–32. doi:10.1080/17434440.2020.1699405

46.

Pechlivanis

Kiriyanthan

Engelhardt

, et al. Percutaneous placement of pedicle screws in the lumbar spine using a bone mounted miniature robotic system: first experiences and accuracy of screw placement. Spine (Phila Pa 1976). 2009;34(4):392–398. doi:10.1097/BRS.0b013e318191ed32

47.

Feng

Tian

Sun

Liu

Wei

. Effect of robot-assisted surgery on lumbar pedicle screw internal fixation in patients with osteoporosis. World Neurosurg. 2019;125:e1057–e1062. doi:10.1016/j.wneu.2019.01.243

48.

Euler Angles. http://encyclopediaofmath.org/index.php?title=Euler_angles&oldid=34483 1748; Accessed April 19, 2021

49.

Siegler

Allard

, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion—part I: ankle, hip, and spine. International Society of Biomechanics. J Biomech. 2002;35(4):543–548. doi:10.1016/s0021-9290(01)00222-6

50.

Euler

. Formulae generales pro translatione quacunque corporum rigidorum. Novi Commentari Academiae Scientiarum Imperalis Petropolitanae. 1775;20:189–207.

51.

Euler

. Nova methodus motum corporum rigidorum determinandi. Novi Commentari Academiae Scientiarum Imperalis Petropolitanae. 1775;20:208–238.

52.

Saha

. Denavit and Hartenberg (DH) Parameters. Introduction to Robotics. 2nd ed. McGraw Hill Education; 2014.

53.

White

Panjabi

. Clinical Biomechanics of the Spine. 2nd ed. J. B. Lippincott; 1990.

54.

Panjabi

. Biomechanical evaluation of spinal fixation devices: I. A conceptual framework. Spine (Phila Pa 1976). 1988;13(10):1129–1134. doi:10.1097/00007632-198810000-00013

55.

Panjabi

Abumi

Duranceau

Crisco

. Biomechanical evaluation of spinal fixation devices: II. Stability provided by eight internal fixation devices. Spine (Phila Pa 1976). 1988;13(10):1135–1140. doi:10.1097/00007632-198810000-00014

56.

ISO 5725-1. Accuracy (Trueness and Precision) of Measurement Methods and Results. Part 1: General Principles and Definitions. 1994.

57.

Elfring

de la Fuente

Radermacher

. Assessment of optical localizer accuracy for computer aided surgery systems. Comput Aided Surg. 2010;15(1-3):1–12. doi:10.3109/10929081003647239

58.

Haidegger

Kazanzides

Rudas

Benyó

. The Importance of Accuracy Measurement Standards for Computer-Integrated Interventional Systems. IEEE; 2010:1–6.

59.

Corazza

Mündermann

Gambaretto

Ferrigno

Andriacchi

. Markerless motion capture through visual hull, articulated ICP and subject specific model generation. Int J Comp Vision. 2010;87(1-2):156–169. doi:10.1007/s11263-009-0284-3

60.

Simon

O’Toole

Blackwell

Morgan

DiGioia

Kanade

. Accuracy validation in image-guided orthopaedic surgery. In: Proceedings of 2nd International Symposium on Medical Robotics and Computer Assisted Surgery. The Robotics Institute Carnegie Mellon University; 1995:185–192.

61.

Sielhorst

Bauer

Wenisch

Klinker

Navab

. Online estimation of the target registration error for n-ocular optical tracking systems. Med Image Comput Comput Assist Interv. 2007;10(Pt 2):652–659. doi:10.1007/978-3-540-75759-7_79

62.

Koivukangas

Katisko

Koivukangas

. Technical accuracy of optical and the electromagnetic tracking systems. Springerplus. 2013;2(1):90. doi:10.1186/2193-1801-2-90

63.

Crouch

Kehl

Krist

. OPTOTRAK: at last a system with resolution of 10 μm. In: Proceedings Volume 1356, Image-Based Motion Measurement. Image-Based Motion Measurement; 1990.

64.

Khadem

Yeh

Sadeghi-Tehrani

, et al. Comparative tracking error analysis of five different optical tracking systems. Comput Aided Surg. 2000;5(2):98–107. doi:10.1002/1097-0150(2000)5:2<98:: AID-IGS4>3.0.CO;2-H

65.

ASTM E177-20. Standard Practice for Use of the Terms Precision and Bias in ASTM Test Methods. ASTM International; 2002.

66.

Maletsky

Sun

Morton

. Accuracy of an optical active-marker system to track the relative motion of rigid bodies. J Biomech. 2007;40(3):682–685. doi:10.1016/j.jbiomech.2006.01.017

67.

Grunert

Darabi

Espinosa

Filippi

. Computer-aided navigation in neurosurgery. Neurosurg Rev. 2003;26(2):73–99. Discussion 100-1. doi:10.1007/s10143-003-0262-0

68.

Simoes

Raposo

Barreto

Edwards

Stoyanov

. Visual tracking vs optical tracking in computer-assisted intervention. IEEE Trans Biomed Eng. 2018.

69.

Chassat

Lavallée

. Experimental protocol of accuracy evaluation of 6-D localizers for computer-integrated surgery: application to four optical localizers. In: Wells

Colchester

Delp

, eds. Medical Image Computing and Computer-Assisted Intervention—MICCAI’98. 1998:277–284. Berlin, Heidelberg. doi:10.1007/bfb0056211

70.

Citak

Kendoff

Wanich

, et al. The influence of distance on registration in ISO-C-3D navigation: a source of error in ISO-C-3D navigation. Technol Health Care. 2006;14(6):473–478.

71.

Fitzpatrick

West

Maurer

. Predicting error in rigid-body point-based registration. IEEE Trans Med Imaging. 1998;17(5):694–702. doi:10.1109/42.736021

72.

Wiles

Thompson

Frantz

. Accuracy Assessment and Interpretation for Optical Tracking Systems. SPIE; 2004:421.

73.

Fitzpatrick

West

. The distribution of target registration error in rigid-body point-based registration. IEEE Trans Med Imaging. 2001;20(9):917–927. doi:10.1109/42.952729

74.

Windolf

Götzen

Morlock

. Systematic accuracy and precision analysis of video motion capturing systems—exemplified on the vicon-460 system. J Biomech. 2008-08-01. 2008;41(12):2776–2780. doi:10.1016/j.jbiomech.2008.06.024

75.

States

Pappas

. Precision and repeatability of the optotrak 3020 motion measurement system. J Med Eng Technol. 2006;30(1):11–16. doi:10.1080/03091900512331304556

76.

Min

Zhu

Meng

MQH

. Accuracy Assessment of an N-ocular Motion Capture System for Surgical Tool Tip Tracking Using Pivot Calibration. IEEE; 2016.

77.

Zhang

Sun

Yuan

Shen

Wang

. A novel method of computer aided orthognathic surgery using individual CAD/CAM templates: a combination of osteotomy and repositioning guides. Br J Oral Maxillofac Surg. 2013;51(8):e239–e244. doi:10.1016/j.bjoms.2013.03.007

78.

Marinetto

Garcia-Mato

Garcia

Martinez

Desco

Pascau

. Multicamera optical tracker assessment for computer aided surgery applications. IEEE Access. 2018;6:64359–64370. doi:10.1109/access.2018.2878323

79.

Gédet

Thistlethwaite

Ferguson

. Minimizing errors during in vitro testing of multisegmental spine specimens: considerations for component selection and kinematic measurement. J Biomech. 2007;40(8):1881–1885. doi:10.1016/j.jbiomech.2006.07.024

80.

Citak

Kendoff

Wanich

, et al. The influence of metal artifacts on navigation and the reduction of artifacts by the use of polyether-ether-ketone. Comput Aided Surg. 2008;13(4):233–239. doi:10.3109/10929080802215292

81.

Stancic

Grujic Supuk

Panjkota

. Design, development and evaluation of optical motion-tracking system based on active white light markers. IET Sci Meas Technol. 2013;7(4):206–214. doi:10.1049/iet-smt.2012.0157

82.

Brinker

Arango

Kaminsky

, et al. An experimental approach to image guided skull base surgery employing a microscope-based neuronavigation system. Acta Neurochir (Wien). 1998;140(9):883–889. doi:10.1007/s007010050189

83.

Dorward

Alberti

Palmer

Kitchen

Thomas

DGT

. Accuracy of true frameless stereotaxy: in vivo measurement and laboratory phantom studies. J Neurosurg. 1999;90(1):160–168. doi:10.3171/jns.1999.90.1.0160

84.

Maurer

Maciunas

Fitzpatrick

. Registration of head CT images to physical space using a weighted combination of points and surfaces. IEEE Trans Med Imaging. 1998;17(5):753–761. doi:10.1109/42.736031

85.

Wilke

Fischer

Jeanneret

Claes

Magerl

. In-vivo-Messung der dreidimensionalen bewegung des iliosakralgelenks. Zeitschrift Fur Orthopadie Und Ihre Grenzgebiet. 2008;135(6):550–556.

86.

Bowden

Guerin

Villarraga

Patwardhan

Ochoa

. Quality of motion considerations in numerical analysis of motion restoring implants of the spine. Clin Biomech (Bristol, Avon). 2008;23(5):536–544. doi:10.1016/j.clinbiomech.2007.12.010

87.

Ilharreborde

Zhao

Boumediene

Gay

Berglund

. A dynamic method for in vitro multisegment spine testing. Orthop Traumatol Surg Res. 2010;96(4):456–461. doi:10.1016/j.otsr.2010.01.006

88.

Jeong

Leasure

Park

. Assessment of biomechanical changes after sacroiliac joint fusion by application of the 3-dimensional motion analysis technique. World Neurosurg. 2018;117:e538–e543. doi:10.1016/j.wneu.2018.06.072

89.

Osterhoff

Dodd

Unno

, et al. Cement augmentation in sacroiliac screw fixation offers modest biomechanical advantages in a cadaver model. Clin Orthop Relat Res. 2016;474(11):2522–2530. doi:10.1007/s11999-016-4934-9

90.

Osterhoff

Tiziani

Hafner

Ferguson

Simmen

Werner

. Symphyseal internal rod fixation versus standard plate fixation for open book pelvic ring injuries: a biomechanical study. Eur J Trauma Emerg Surg. 2016;42(2):197–202. doi:10.1007/s00068-015-0529-5

91.

Hammer

Scholze

Kibsgård

, et al. Physiological in vitro sacroiliac joint motion: a study on three-dimensional posterior pelvic ring kinematics. J Anat. 2019;234(3):346–358. doi:10.1111/joa.12924

92.

Baria

Lindsey

Milne

Kaimrajh

Latta

. Effects of lumbosacral arthrodesis on the biomechanics of the sacroiliac joint. JB JS Open Access. 2020;5(1):e0034. doi:10.2106/JBJS.OA.19.00034

93.

Devito

Kaplan

Dietl

, et al. Clinical acceptance and accuracy assessment of spinal implants guided with SpineAssist surgical robot: retrospective study. Spine (Phila Pa 1976). 2010;35(24):2109–2115. doi:10.1097/BRS.0b013e3181d323ab

94.

Helm

Teichman

Hartmann

Simon

. Spinal navigation and imaging: history, trends, and future. IEEE Trans Med Imaging. 2015;34(8):1738–1746. doi:10.1109/TMI.2015.2391200

95.

Ringel

Stüer

Reinke

, et al. Accuracy of robot-assisted placement of lumbar and sacral pedicle screws: a prospective randomized comparison to conventional freehand screw implantation. Spine (Phila Pa 1976). 2012;37(8):E496–E501. doi:10.1097/BRS.0b013e31824b7767

96.

Fan

Liu

, et al. Accuracy of pedicle screw placement comparing robot-assisted technology and the free-hand with fluoroscopy-guided method in spine surgery: an updated meta-analysis. Medicine (Baltimore). 2018;97(22):e10970. doi:10.1097/md.0000000000010970

97.

Fan

Liu

, et al. Improved accuracy of cervical spinal surgery with robot-assisted screw insertion: a prospective, randomized, controlled study. Spine (Phila Pa 1976). 2020;45(5):285–291. doi:10.1097/BRS.0000000000003258

98.

Huang

Tetreault

Vaishnav

York

Staub

. The current state of navigation in robotic spine surgery. Ann Transl Med. 2021;9(1):86. doi:10.21037/atm-2020-ioi-07

99.

Kalidindi

KKV

Sharma

Jagadeesh

Sath

Chhabra

. Robotic spine surgery: a review of the present status. J Med Eng Technol. 2020;44(7):431–437. doi:10.1080/03091902.2020.1799098

100.

Elswick

Strong

Joseph

Saadeh

Oppenlander

Park

. Robotic-assisted spinal surgery: current generation instrumentation and new applications. Neurosurg Clin N Am. 2020;31(1):103–110. doi:10.1016/j.nec.2019.08.012

101.

Overley

Cho

Mehta

Arnold

. Navigation and robotics in spinal surgery: where are we now? Neurosurgery. 2017;80(3S):S86–S99. doi:10.1093/neuros/nyw077

102.

. Spinal navigation with intra-operative 3D-imaging modality in lumbar pedicle screw fixation. Zhonghua Yi Xue Za Zhi. 2008;88(27):1905–1908.

103.

Bajwa

Toy

Ahn

. Pedicle morphometry of upper thoracic vertebrae: an anatomic study of 503 cadaveric specimens. Spine (Phila Pa 1976). 2014;39(20):E1201–E1209. doi:10.1097/BRS.0000000000000505

104.

Chen

Margalit

Peng

Qiu

Qian

. Robot-assisted vs freehand pedicle screw fixation in spine surgery—a systematic review and a meta-analysis of comparative studies. Int J Med Robot. 2018;14(3):e1892. doi:10.1002/rcs.1892

105.

Vardiman

Wallace

Booher

, et al.

Does the accuracy of pedicle screw placement differ between the attending surgeon and resident in navigated robotic-assisted minimally invasive spine surgery?

J Robot Surg. 2020;14(4):567–572. doi:10.1007/s11701-019-01019-9

106.

Vardiman

Wallace

Crawford

Riggleman

Ahrendtsen

Ledonio

. Pedicle screw accuracy in clinical utilization of minimally invasive navigated robot-assisted spine surgery. J Robot Surg. 2020;14(3):409–413. doi:10.1007/s11701-019-00994-3

107.

Keric

Doenitz

Haj

, et al. Evaluation of robot-guided minimally invasive implantation of 2067 pedicle screws. Neurosurg Focus. 2017;42(5):E11. doi:10.3171/2017.2.FOCUS16552

108.

Chen

Xiao

Chen

, et al. Results of using robotic-assisted navigational system in pedicle screw placement. PLoS One. 2019;14(8):e0220851. doi:10.1371/journal.pone.0220851

109.

Chen

Xiao

Chen

, et al. A spine robotic-assisted navigation system for pedicle screw placement. J Vis Exp. 2020;(159). doi:10.3791/60924

110.

Chen

Zhang

Zhan

Gan

Sun

. Research on the accuracy of three-dimensional localization and navigation in robot-assisted spine surgery. Int J Med Robot. 2020;16(2):e2071. doi:10.1002/rcs.2071

111.

Togawa

Kayanja

Reinhardt

, et al. Bone-mounted miniature robotic guidance for pedicle screw and translaminar facet screw placement: part 2—evaluation of system accuracy. Neurosurgery. 2007;60(2 Suppl 1):ONS129–139. Discussion ONS139. doi:10.1227/01.NEU.0000249257.16912.AA

112.

Macke

Woo

Varich

. Accuracy of robot-assisted pedicle screw placement for adolescent idiopathic scoliosis in the pediatric population. J Robot Surg. 2016;10(2):145–150. doi:10.1007/s11701-016-0587-7

113.

Tong

Liu

, et al. Robot-assisted technique vs conventional freehand technique in spine surgery: a meta-analysis. Int J Clin Pract. 2020;75(5):e13964. doi:10.1111/ijcp.13964

114.

Huntsman

Riggleman

Ahrendtsen

Ledonio

. Navigated robot-guided pedicle screws placed successfully in single-position lateral lumbar interbody fusion. J Robot Surg. 2020;14(4):643–647. doi:10.1007/s11701-019-01034-w

115.

Johnson

. Imaging, navigation, and robotics in spine surgery. Spine (Phila Pa 1976). 2016;41(Suppl 7):S32. doi:10.1097/BRS.0000000000001437

116.

Kochanski

Lombardi

Laratta

Lehman

O’Toole

. Image-guided navigation and robotics in spine surgery. Neurosurgery. 2019;84(6):1179–1189. doi:10.1093/neuros/nyy630

117.

Kotani

Abumi

Ito

Minami

. Improved accuracy of computer-assisted cervical pedicle screw insertion. J Neurosurg. 2003;99(3 Suppl):257–263. doi:10.3171/spi.2003.99.3.0257

118.

Tian

Shi

, et al. Robot-assisted versus fluoroscopy-assisted cortical bone trajectory screw instrumentation in lumbar spinal surgery: a matched-cohort comparison. World Neurosurg. 2018;120:e745–e751. doi:10.1016/j.wneu.2018.08.157

119.

Sukovich

Brink-Danan

Hardenbrook

. Miniature robotic guidance for pedicle screw placement in posterior spinal fusion: early clinical experience with the SpineAssist. Int J Med Robot. 2006;2(2):114–122. doi:10.1002/rcs.86

120.

Jiang

Azad

Crawford

Bydon

Theodore

Robotic spine surgery: current state in minimally invasive surgery. Global Spine J. 2020;10(2 Suppl):34S–40S. doi:10.1177/2192568219878131

121.

Wallace

Vardiman

Booher

, et al. Navigated robotic assistance improves pedicle screw accuracy in minimally invasive surgery of the lumbosacral spine: 600 pedicle screws in a single institution. Int J Med Robot. 2020;16(1):e2054. doi:10.1002/rcs.2054

122.

West

Fitzpatrick

Toms

Maurer

Maciunas

. Fiducial point placement and the accuracy of point-based, rigid body registration. Neurosurgery. 2001;48(4):810–816. Discussion 816-7. doi:10.1097/00006123-200104000-00023

123.

West

Maurer

. Designing optically tracked instruments for image-guided surgery. IEEE Trans Med Imaging. 2004;23(5):533–545. doi:10.1109/tmi.2004.825614

124.

Gertzbein

Robbins

. Accuracy of pedicular screw placement in vivo. Spine (Phila Pa 1976). 1990;15(1):11–14. doi:10.1097/00007632-199001000-00004

125.

Verma

Santoshi

Jain

, et al. Thoracic pedicle morphometry of dry vertebral columns in relation to trans-pedicular fixation: a cross-sectional study from central India. Cureus. 2020;12(5):e8148. doi:10.7759/cureus.8148

126.

Holly

. Image-guided spinal surgery. Int J Med Robot. 2006;2(1):7–15. doi:10.1002/rcs.69

127.

Gupta

Kapoor

Sharma

Kochhar

Garg

. Morphometry of typical cervical vertebrae on dry bones and CT scan and its implications in transpedicular screw placement surgery. Surg Radiol Anat. 2013;35(3):181–189. doi:10.1007/s00276-012-1013-0

128.

Herrero

Luis do Nascimento

Maranho

DAC

, et al. Cervical pedicle morphometry in a Latin American population: a Brazilian study. Medicine (Baltimore). 2016;95(25):e3947. doi:10.1097/MD.0000000000003947

129.

Attar

Ugur

Tekdemir

Egemen

Genc

. Lumbar pedicle: surgical anatomic evaluation and relationships. Eur Spine J. 2001;10(1):10–15. doi:10.1007/s005860000198

130.

Avrunin

Alkhorayef

Farouk Ismail Saied

Tymkovych

. The surgical navigation system with optical position determination technology and sources of errors. J Med Imaging Health Inform. 2015;5(4):689–696. doi:10.1166/jmihi.2015.1444

131.

Garg

Mehta

Malhotra

. Robotic spine surgery: ushering in a new era. J Clin Orthop Trauma. 2020;11(5):753–760. doi:10.1016/j.jcot.2020.04.034

132.

Ohnmeiss

Lieberman

. Robotic-assisted pedicle screw placement: lessons learned from the first 102 patients. Eur Spine J. 2013;22(3):661–666. doi:10.1007/s00586-012-2499-1

133.

Lieberman

. What is the learning curve for robotic-assisted pedicle screw placement in spine surgery? Clin Orthop Relat Res. 2014;472(6):1839–1844. doi:10.1007/s11999-013-3291-1

134.

Abbasi

Grzeszczuk

Chin

, et al. Clinical fluoroscopic fiducial-based registration of the vertebral body in spinal neuronavigation. Stud Health Technol Inform. 2001;81:1–7.

135.

Hüfner

Geerling

Oldag

, et al. Accuracy study of computer-assisted drilling: the effect of bone density, drill bit characteristics, and use of a mechanical guide. J Orthop Trauma. 2005;19(5):317–322.

136.

Messmer

Gross

Suhm

Regazzoni

Jacob

Huegli

. Modality-based navigation. Injury. 2004;35(Suppl 1):S–A24–A29. doi:10.1016/j.injury.2004.05.007

137.

Rahmathulla

Nottmeier

Pirris

Deen

Pichelmann

. Intraoperative image-guided spinal navigation: technical pitfalls and their avoidance. Neurosurg Focus. 2014;36(3):E3. doi:10.3171/2014.1.focus13516

Comparative Analysis of Optoelectronic Accuracy in the Laboratory Setting Versus Clinical Operative Environment: A Systematic Review

Abstract

Study Design:

Objectives:

Methods:

Results:

Conclusions:

Keywords

Introduction

Methods

Results

Three-Dimensional Cartesian Rigid Body Transformations

Optoelectronic Measurement Systems

Intrinsic (Technical) Accuracy

Registration Accuracy and Target Tracking

Application Accuracy—Musculoskeletal Kinematics Laboratory Platform

Application Accuracy—Clinical Operative Platform

Basis for Disparity in Optoelectronic Accuracy

Discussion

Footnotes

Acknowledgments

Declaration of Conflicting Interests

Funding

ORCID iD

References