Dynamic assessment of spine movement patterns using an RGB-D camera and deep learning.

In clinical practice, functional limitations in patients with low back pain are subjectively assessed, potentially leading to misdiagnosis and prolonged pain. This paper proposes an objective deep learning (DL) markerless motion capture system that uses a red-green-blue-depth (RGB-D) camera to measure the kinematics of the spine during flexion-extension (FE) through: 1) the development and validation of a DL semantic segmentation algorithm that segments the back into four anatomical classes and 2) the development and validation of a framework that uses these segmentations to measure spine kinematics during FE. Twenty participants performed ten cycles of FE with drawn-on point markers while being recorded with an RGB-D camera. Five of these participants also performed an additional trial where they were recorded with an optical motion capture (OPT) system. The DL algorithm was trained to segment the back and pelvis into four anatomical classes: upper back, lower back, spine, and pelvis. A kinematic framework was then developed to refine these segmentations into upper spine, lower spine, and pelvis masks, which were used to measure spine kinematics after obtaining 3D global coordinates of the mask corners. The segmentation algorithm achieved high accuracy, and the root mean square error (RMSE) between ground truth and predicted lumbar kinematics was < 4°. When comparing markerless and OPT kinematics, RMSE values were < 6°. This work demonstrates the feasibility of using markerless motion capture to assess FE spine movement in clinical settings. Future work will expand the studied movement directions and test on different demographics.

Investigating concurrent validity of inertial sensors to evaluate multiplanar spine movement.

Inertial measurement units (IMUs) offer a portable and inexpensive alternative to traditional optical motion capture systems, and have potential to support clinical diagnosis and treatment of low back pain; however, due to a lack of confidence regarding the validity of IMU-derived metrics, their uptake and acceptance remain a challenge. The objective of this work was to assess the concurrent validity of the Xsens DOT IMUs for tracking multiplanar spine movement, and to evaluate concurrent validity and reliability for estimating clinically relevant metrics relative to gold-standard optical motion capture equipment. Ten healthy controls performed spine range of motion (ROM) tasks, while data were simultaneously tracked from IMUs and optical marker clusters placed over the C7, T12, and S1 vertebrae. Root mean square error (RMSE), mean absolute error (MAE), and intraclass correlation coefficients (ICC2,1) were calculated to assess validity and reliability of absolute (abs; C7, T12, and S1 sensors) and relative joint (rel; intersegmental thoracic, lumbar, and total) motion. Overall RMSEabs = 1.33°, MAEabs = 0.74° ± 0.69, and ICC2,1,abs = 0.953 across all movements, sensors, and planes. Results were slightly better for uniplanar movements when evaluating the primary rotation axis (prim) absolute ROM (MAEabs,prim = 0.56° ± 0.49; ICC2,1,abs,prim = 0.999). Similarly, when evaluating relative intersegmental motion, overall RMSErel = 2.39°, MAErel = 1.10° ± 0.96, and ICC2,1,rel = 0.950, and relative primary rotation axis achieved MAErel,prim = 0.87° ± 0.77, and ICC2,1,rel,prim = 0.994. Findings from this study suggest that these IMUs can be considered valid for tracking multiplanar spine movement, and may be used to objectively assess spine movement and neuromuscular control in clinics.

An Enhanced Spine Model Validated for Simulating Dynamic Lifting Tasks in OpenSim.

A fully articulated thoracolumbar spine model had been previously developed in OpenSim and had been extensively validated against experimental data during various static tasks. In the present study, we enhanced this detailed musculoskeletal model by adding the role of passive structures and adding kinematic constraints to make it suitable for dynamic tasks. We validated the spinal forces estimated by this enhanced model during nine dynamic lifting/lowering tasks. Moreover, we recently developed and evaluated five approaches in OpenSim to model the external loads applied to the hands during lifting/lowering tasks, and in the present study, we assessed which approach results in more accurate spinal forces. Regardless of the external load modeling approach, the maximum forces predicted by our enhanced spine model across all tasks, as well as the pattern of estimated spinal forces within each task, showed strong correlations (r-values and cross-correlation coefficients > 0.9) with experimental data. Given the biofidelity of our enhanced model, its accessibility via the open-source OpenSim software, and the extent to which this model has been validated, we recommend it for applications requiring estimation of spinal forces during lifting/lowering tasks using multibody-based models and inverse dynamic analyses.

Smartwatch-Based Prediction of Single-Stride and Stride-to-Stride Gait Outcomes Using Regression-Based Machine Learning.

Spatiotemporal variability during gait is linked to fall risk and could be monitored using wearable sensors. Although many users prefer wrist-worn sensors, most applications position at other sites. We developed and evaluated an application using a consumer-grade smartwatch inertial measurement unit (IMU). Young adults (n = 41) completed seven-minute conditions of treadmill gait at three speeds. Single-stride outcomes (stride time, length, width, and speed) and spatiotemporal variability (coefficient of variation of each single-stride outcome) were recorded using an optoelectronic system, while 232 single- and multi-stride IMU metrics were recorded using an Apple Watch Series 5. These metrics were input to train linear, ridge, support vector machine (SVM), random forest, and extreme gradient boosting (xGB) models of each spatiotemporal outcome. We conducted Model × Condition ANOVAs to explore model sensitivity to speed-related responses. xGB models were best for single-stride outcomes [relative mean absolute error (% error): 7-11%; intraclass correlation coefficient (ICC2,1) 0.60-0.86], and SVM models were best for spatiotemporal variability (% error: 18-22%; ICC2,1 = 0.47-0.64). Spatiotemporal changes with speed were captured by these models (Condition: p < 0.00625). Results support the feasibility of monitoring single-stride and multi-stride spatiotemporal parameters using a smartwatch IMU and machine learning.

Predicting vertical and shear ground reaction forces during walking and jogging using wearable plantar pressure insoles.

BACKGROUND: The development of plantar pressure insoles has made them a potential replacement for force plates. These wearable devices can measure multiple steps and might be used outside of the lab environment for rehabilitation and evaluation of sport performance. However, they can only measure the vertical force which does not completely represent the vertical ground reaction force. In addition, they are not able to measure shear forces which play an import role in the dynamic performance of individuals. Indirect approaches might be implemented to improve the accuracy of the force estimated by plantar pressure systems. RESEARCH QUESTION: The aim of this study was to predict the vertical and shear components of ground reaction force from plantar pressure data using recurrent neural networks. METHODS: Ground reaction force and plantar pressure data were collected from 16 healthy individuals during 10 trials of walking and five trials of jogging using Bertec force plates at 1000 Hz and FScan plantar pressure insoles at 100 Hz. A long short-term memory neural network was built to consider the time dependency of pressure and force data in predictions. The data were split into three subsets of train, to train the model, evaluate, to optimize the model hyperparameters, and test sets, to assess the accuracy of the model predictions. RESULTS: The results of this study showed that our long short-term memory model could accurately predict the shear and vertical force components during walking and jogging. The predictions were more accurate during walking compared to jogging. In addition, the predictions of mediolateral force had higher error and lower correlation compared to vertical and anteroposterior components. SIGNIFICANCE: The long short-term memory model developed in this study may be an acceptable option for accurate estimation of ground reaction force during outdoor activities which can have significant impacts in rehabilitation, sport performance, and gaming.

Three-Dimensional Motion Capture Data of a Movement Screen from 183 Athletes.

Movement screens are widely used to identify aberrant movement patterns in hopes of decreasing risk of injury, identifying talent, and/or improving performance. Motion capture data can provide quantitative, objective feedback regarding movement patterns. The dataset contains three-dimensional (3D) motion capture data of 183 athletes performing mobility tests (ankle, back bend, crossover adduction, crossover rotation, elbows, head, hip turn, scorpion, shoulder abduction, shoulder azimuth, shoulder rotation, side bends, side lunges and trunk rotation) and stability tests (drop jump, hop down, L-cut, lunge, rotary stability, step down and T-balance) bilaterally (where applicable), the athletes' injury history, and demographics. All data were collected at 120 Hz or 480 Hz using an 8-camera Raptor-E motion capture system with 45 passive reflective markers. A total of 5,493 trials were pre-processed and included in .c3d and .mat formats. This dataset will enable researchers and end users to explore movement patterns of athletes of varying demographics from different sports and competition levels; develop objective movement assessment tools; and gain new insights into the relationships between movement patterns and injury.

Joint behaviour during arm swing changes with gait speed and predicts spatiotemporal variability and dynamic stability in healthy young adults.

BACKGROUND: Arm swing is linked to gait stability. How this is accomplished is unclear as most investigations artificially manipulate arm swing amplitude and examine average patterns. Biomechanical evaluation of stride-to-stride upper limb behaviour across a range of gait speeds, where the arm swings as preferred, could clarify this link. RESEARCH QUESTION: How do stride-to-stride arm swing behaviours change with gait speed and relate to stride-to-stride gait fluctuations? METHODS: Young adults (n = 45, 25 females) completed treadmill gait at preferred, slow (70% of preferred), and fast speed (130% of preferred) while full-body kinematics were acquired with optoelectronic motion capture. Arm swing behaviour was quantified by shoulder, elbow, and wrist joint angle amplitude (range of motion [ROM]) and motor variability (e.g. mean standard deviation [meanSD], local divergence exponent [λmax]). Stride-to-stride gait fluctuation was quantified by spatiotemporal variability (e.g. stride time CV) and dynamic stability (i.e. trunk local dynamic stability [trunk λmax], centre-of-mass smoothness [COM HR]). Repeated measures ANOVAs tested for speed effects and step-wise linear regressions identified arm swing-based predictors of stride-to-stride gait fluctuation. RESULTS: Speed decreased spatiotemporal variability and increased trunk λmax and COM HR in the anteroposterior and vertical axes. Adjustments in gait fluctuations occurred with increased upper limb ROM, particularly for elbow flexion, and increased meanSD and λmax of shoulder, elbow, and wrist angles. Models of upper limb measures predicted 49.9-55.5% of spatiotemporal variability and 17.7-46.4% of dynamic stability. For dynamic stability, wrist angle features were the best and most common independent predictors. SIGNIFICANCE: Findings highlight that all upper limb joints, and not solely the shoulder, underlie changes in arm swing amplitude, and that arm swing strategies pair with the trunk and contrast with centre-of-mass and stride strategies. Findings suggest that young adults search for flexible arm swing motor strategies to help optimize stride consistency and gait smoothness.

Morphable models of the lumbar spine to vary geometry based on pathology, demographics, and anatomical measurements.

The shape of the lumbar spine influences its function and dysfunction. Yet examining the influence of geometric differences associated with pathology or demographics on lumbar biomechanics is challenging in vivo where these effects cannot be isolated, and the use of simple anatomical measurements does not fully capture the complex three-dimensional geometry. The goal of this work was to develop and share morphable models of the lumbar spine that allow geometry to be varied according to pathology, demographics, or anatomical measurements. Partial least squares regression was used to generate statistical shape models that quantify geometric differences associated with pathology, demographics, and anatomical measurements from the lumbar spines of 87 patients. To determine if the morphable models detected meaningful geometric differences, the ability of the morphable models to classify spines was compared with models generated from random labels. The models for disc herniation (p < 0.04), spondylolisthesis (p < 0.001), and sex (p < 0.01) all performed significantly better than the random models. Age was predicted with a root mean square error of 14.1 years using the age-based model. The morphable models for anatomical measurements were able to produce instances with root mean square errors less than 0.8°, 0.3 cm2, and 0.7 mm between desired and resulting measurements. This method can be used to produce morphable models that enable further analysis of the relationship among shape, pathology, demographics, and function through computational simulations. The morphable models and code are available at https://github.com/aclouthier/morphable-lumbar-model.

Evaluation of spinal force normalization techniques.

Division normalization is commonly used in biomechanics studies to remove the effect of anthropometric differences (e.g., body weight) on kinetic variables, facilitating comparison across a population. In spine biomechanics, spinal forces are commonly divided by the body weight or the intervertebral load during a standing posture. However, it has been suggested that offset and power curve normalization are more appropriate than division normalization for normalizing kinetic variables such as ground reaction forces during walking and running. The present study investigated, for the first time, the effectiveness of four techniques for normalizing spinal forces to remove the effect of body weight. Spinal forces at all lumbar levels were estimated using a detailed OpenSim musculoskeletal model of the spine for 11 scaled models (50-100 kg) and during 13 trunk flexion tasks. Pearson correlations of raw and normalized forces against body weight were used to assess the effectiveness of each normalization technique. Body weight and standing division normalization could only successfully normalize L4L5 spinal forces in three tasks, and L5S1 loads in five and three tasks, respectively; however, offset and power curve normalization techniques were successful across all lumbar spine levels and all tasks. Offset normalization successfully removed the effect of body weight and maintained the influence of flexion angle on spinal forces. Thus, we recommend offset normalization to account for anthropometric differences in studies of spinal forces.

Exploring the relationship between kinematic variability and fatigue development during repetitive lifting.

To investigate the variability-fatigue and repeaters-replacers hypotheses, motor variability (MV) and indicators of fatigue were assessed during repetitive lifting. Eighteen participants performed sequential repetitive bouts of lifting divided into a short bout, and three phases of a prolonged bout until volitional fatigue (or until a 1-h time limit). Whole-body kinematics were collected to calculate variability in three-dimensional joint angles and in continuous relative phase (CRP) of sagittal joint angle couplings, which were summed for the upper and lower body, and whole-body. Excellent individual consistency (ICC = 0.95-0.97) was demonstrated across lifting bouts as fatigue developed. Therefore, strong evidence was obtained for MV as an individual trait in support of the repeaters-replacers hypothesis. Associations were found for endurance and baseline effort with lower body variability, while no associations were found for rate of fatigue. Thus, some support was found for the variability-fatigue hypothesis which suggests that repeaters are less fatigue-resistant than replacers.

Exploring the role of task on kinematic variability and assessing consistency in individual responses across repetitive manual tasks.

To gain a greater understanding of motor variability (MV) as an individual trait, the effect of task type on MV and individual consistency in MV across three tasks was investigated. Twenty participants performed repetitive carrying, lifting, and simulated sawing tasks. MV was assessed using the linear measure of mean point-by-point standard deviation in three-dimensional upper body joint angles. Task type affected MV, where carrying showed higher MV compared to sawing (23-29%) and lifting (12-19%). Furthermore, MV was higher in lifting compared to sawing (12-25%). Poor to moderate individual consistency (ICC = 0.42-0.63) was found across tasks. Task type determined MV and only some support for MV as an individual trait across tasks was found. Based on this work, differences in degrees of freedom afforded by the task influence the opportunity to exploit MV, and possibly individual consistency in MV magnitude is specific to the degrees of freedom afforded by the task. Practitioner summary: In repetitive tasks, movement variability has been proposed as an individual characteristic independent of task characteristics, where repeaters show consistently low variability, while replacers show consistently high variability. In the current study, only moderate support was demonstrated for variability as a consistent individual characteristic across different manual tasks.AbbreviationMV: Motor variability; WRMSDs: Work-related musculoskeletal disorders; DOF: Degrees of freedom; meanSD: Mean standard deviation; SD: Standard deviation; H: Handle (of simulated sawing setup); T: Track (of simulated sawing setup); F: Frame (of simulated sawing setup); ICC: Intraclass correlation; UE: Upper extremity; MMH: Manual material handling; EMG: Electromyography.

The effect of vertebral body tethering on spine range of motion in adolescent idiopathic scoliosis: a pilot study.

PURPOSE: Posterior spinal fusion and instrumentation (PSF) and vertebral body tethering (VBT) are corrective surgical techniques used in treating adolescent idiopathic scoliosis (AIS). Comparing the preservation of spine range of motion (ROM) following PSF and VBT for treatment of AIS has yet to be explored. The purpose of this work was to retrospectively compare global spine ROM in adolescents (9-18 years of age) without spine deformity, adolescents with untreated AIS, adolescents having undergone PSF, and adolescents having undergone VBT to gain insight on the effect of VBT on spine motion. METHODS: Twenty participants were recruited into four groups including Control (n = 6), untreated AIS (n = 5), post-operative PSF (n = 4) and post-operative VBT (n = 5). Three-dimensional kinematics of the spine were collected and analyzed using an intersegmental spine model during constrained forward flexion, right-left lateral bending, and right-left axial twist movements. RESULTS: The PSF group displayed significantly lower spine ROM than the two non-operative groups during thoracic and total left axial twist (p ≤ 0.048), whereas thoracic and total ROM during right-left lateral bending is almost equally lower in the PSF (p ≤ 0.03) and VBT (p ≤ 0.01) groups when compared to the Control and AIS groups. CONCLUSION: These results suggest some preservation of spine motion in the transverse plane following VBT. This study provides initial evidence of some potential preservation of spine ROM following VBT; however, further prospective investigation of VBT is needed to assess and confirm these hypotheses.

Comparison of machine learning classifiers for differentiating level and sport using movement data.

The purposes of this study were to determine if 1) recurrent neural networks designed for multivariate, time-series analyses outperform traditional linear and non-linear machine learning classifiers when classifying athletes based on competition level and sport played, and 2) athletes of different sports move differently during non-sport-specific movement screens. Optical-based kinematic data from 542 athletes were used as input data for nine different machine learning algorithms to classify athletes based on competition level and sport played. For the traditional machine learning classifiers, principal component analysis and feature selection were used to reduce the data dimensionality and to determine the best principal components to retain. Across tasks, recurrent neural networks and linear machine learning classifiers tended to outperform the non-linear machine learning classifiers. For all tasks, reservoir computing took the least amount of time to train. Across tasks, reservoir computing had one of the highest classification rates and took the least amount of time to train; however, interpreting the results is more difficult compared to linear classifiers. In addition, athletes were successfully classified based on sport suggesting that athletes competing in different sports move differently during non-sport specific movements. Therefore, movement assessment screens should incorporate sport-specific scoring criteria.

A data-driven framework for assessing soldier performance, health, and survivability.

Presented is a framework that uses pattern classification methods to incrementally morph whole-body movement patterns to investigate how personal (sex, military experience, and body mass) and load characteristics affect the survivability tradespace: performance, musculoskeletal health, and susceptibility to enemy action. Sixteen civilians and 12 soldiers performed eight military-based movement patterns under three body-borne loads: ∼5.5 kg, ∼22 kg, and ∼38 kg. Our framework reduces dimensionality using principal component analysis and uses linear discriminant analysis to classify groups and morph movement patterns. Our framework produces morphed whole-body movement patterns that emulate previously published changes to the survivability tradespace caused by body-borne loads. Additionally, we identified that personal characteristics can greatly impact the tradespace when carrying heavy body-borne loads. Using our framework, military leaders can make decisions based on objective information for armour procurement, employment of armour, and battlefield performance, which can positively impact operational readiness and increase overall mission success.

The effects of mobile phone use on motor variability patterns during gait.

Mobile phone use affects the dynamics of gait by impairing visual control of the surrounding environment and introducing additional cognitive demands. Although it has been shown that using a mobile phone alters whole-body dynamic stability, no clear information exists on its impacts on motor variability during gait. This study aimed at assessing the impacts of various types of mobile phone use on motor variability during gait; quantified using the short- and long-term Lyapunov Exponent (λS and λL) of lower limb joint angles and muscle activation patterns, as well as the centre of mass position. Fourteen females and Fifteen males (27.72 ± 4.61 years, body mass: 70.24 ± 14.13 Kg, height: 173.31 ± 10.97 cm) walked on a treadmill under six conditions: normal walking, normal walking in low-light, walking while looking at the phone, walking while looking at the phone in low-light, walking and talking on the phone, and walking and listening to music. Variability of the hip (p λS = .015, λL = .043) and pelvis (p λS = .039, λL = .017) joint sagittal angles significantly increased when the participants walked and looked at the phone, either in normal or in low-light conditions. No significant difference was observed in the variability of the centre of mass position and muscle activation patterns. When individuals walk and look at the phone screen, the hip and knee joints are constantly trying to adopt a new angle to regulate and maintain gait stability, which might put an additional strain on the neuromuscular system. To this end, it is recommended not to look at the mobile phone screen while walking, particularly in public places with higher risks of falls.

The role of machine learning in the primary prevention of work-related musculoskeletal disorders: A scoping review.

To determine the applications of machine learning (ML) techniques used for the primary prevention of work-related musculoskeletal disorders (WMSDs), a scoping review was conducted using seven literature databases. Of the 4,639 initial results, 130 primary research studies were deemed relevant for inclusion. Studies were reviewed and classified as a contribution to one of six steps within the primary WMSD prevention research framework by van der Beek et al. (2017). ML techniques provided the greatest contributions to the development of interventions (48 studies), followed by risk factor identification (33 studies), underlying mechanisms (29 studies), incidence of WMSDs (14 studies), evaluation of interventions (6 studies), and implementation of effective interventions (0 studies). Nearly a quarter (23.8%) of all included studies were published in 2020. These findings provide insight into the breadth of ML techniques used for primary WMSD prevention and can help identify areas for future research and development.

Effects of arm swing amplitude and lower limb asymmetry on motor variability patterns during treadmill gait.

Motor variability is a fundamental feature of gait. Altered arm swing and lower limb asymmetry (LLA) may be contributing factors having been shown to affect the magnitude and dynamics of variability in spatiotemporal and trunk motion. However, the effects on lower limb joints remain unclear. Full-body kinematics of 15 healthy young adults were recorded during treadmill walking using the Computer-Assisted Rehabilitation Environment system. Participants completed six trials, combining three arm swing (AS) amplitude (normal, active, held) and two LLA (symmetrical, asymmetrical) conditions. The mean standard deviation (meanSD), maximum Lyapunov exponent (λmax), detrended fluctuation analysis scaling exponent of range of motion (DFAα), and sample entropy (SaEn) were computed for tridimensional trunk, pelvis, and lower limb joint angles, and compared using repeated-measures ANOVAs. Relative to normal AS, active AS increased meanSD of all joint angles, λmax of frontal plane hip and ankle angles, and SaEn of sagittal plane ankle angles. Active AS, however, did not affect λmax or SaEn of trunk or pelvis angles. LLA increased meanSD of sagittal plane joint angles, λmax of Euclidean norm trunk angle and of lower limb joint angles, and SaEn of ankle dorsiflexion/ plantarflexion, but decreased SaEn of tridimensional trunk angles and hip rotation in the slower moving leg. Alterations in lower limb variability with active AS and LLA suggest that young adults actively exploit their lower limb redundancies to maintain gait. This appears to preserve trunk stability and regularity during active AS but not during LLA.

Exploring the role of task constraints on motor variability and assessing consistency in individual responses during repetitive lifting using linear variability of kinematics.

To better understand the assessment of motor variability (MV) in an occupational context, this study determined the role of task constraints on MV and consistency in individual MV responses. Twenty participants performed repetitive lifting under four constraints differing in restriction of foot movement and load weight. MV was assessed for three body regions and for the whole-body using linear variability of three-dimensional joint angles. Foot movement caused significant increases of lower body (11-17%), low back (318-439%) and a reduction in upper body variability (4%), whereas no effects of weight nor interaction of foot restriction and weight were found. Good individual consistency (ICC = 0.71-0.84) was demonstrated across constraints. Even though MV is affected by constraints, this study supports that MV is largely an individual trait independent of constraints. Future work should evaluate if MV remains an individual trait across different tasks, and if MV is confounded by other task constraints.

Validity and Sensitivity of an Inertial Measurement Unit-Driven Biomechanical Model of Motor Variability for Gait.

Motor variability in gait is frequently linked to fall risk, yet field-based biomechanical joint evaluations are scarce. We evaluated the validity and sensitivity of an inertial measurement unit (IMU)-driven biomechanical model of joint angle variability for gait. Fourteen healthy young adults completed seven-minute trials of treadmill gait at several speeds and arm swing amplitudes. Trunk, pelvis, and lower-limb joint kinematics were estimated by IMU- and optoelectronic-based models using OpenSim. We calculated range of motion (ROM), magnitude of variability (meanSD), local dynamic stability (λmax), persistence of ROM fluctuations (DFAα), and regularity (SaEn) of each angle over 200 continuous strides, and evaluated model accuracy (RMSD: root mean square difference), consistency (ICC2,1: intraclass correlation), biases, limits of agreement, and sensitivity to within-participant gait responses (effects of speed and swing). RMSDs of joint angles were 1.7-9.2° (pooled mean of 4.8°), excluding ankle inversion. ICCs were mostly good to excellent in the primary plane of motion for ROM and in all planes for meanSD and λmax, but were poor to moderate for DFAα and SaEn. Modelled speed and swing responses for ROM, meanSD, and λmax were similar. Results suggest that the IMU-driven model is valid and sensitive for field-based assessments of joint angle time series, ROM in the primary plane of motion, magnitude of variability, and local dynamic stability.

Concurrent validity of a custom computer vision algorithm for measuring lumbar spine motion from RGB-D camera depth data.

Using RGB-D cameras as an alternative motion capture device can be advantageous for biomechanical spine motion assessments of movement quality and dysfunction due to their lower cost and complexity. In this study, we evaluated RGB-D camera performance relative to gold-standard optoelectronic motion capture equipment. Twelve healthy young adults (6M, 6F) were recruited to perform repetitive spine flexion-extension, while wearing infrared reflective marker clusters placed over their T10-T12 spinous processes and sacrum, and motion capture data were recorded simultaneously by both systems. Custom computer vision algorithms were developed to extract spine angles from depth data. Root mean square error (RMSE) was calculated for continuous Euler angles, and intraclass correlation coefficients (ICC2,1) were calculated between minimum and maximum angles and range of motion in all movement planes. RMSE was low (RMSE ≤ 2.05°) and reliability was good to excellent (0.849 ≤ ICC2,1 ≤ 0.979) across all movement planes. In conclusion, the proposed algorithm for tracking 3D lumbar spine motion during a sagittal movement task from one RGB-D camera is reliable in comparison to gold-standard motion tracking equipment. Future research will investigate accuracy and validity in a wider variety of movements, and will also investigate the development of novel methods to measure spine motion without using infrared reflective markers.