Motion Capture Markers Research Articles

The Influence of Hydraulic Ankles and Microprocessor-control on the Biomechanics of Trans-tibial Amputees During Quiet Standing on a 5° Slope.

Lower limb amputees have a high incidence of comorbidities, such as osteoarthritis, which are believed to be caused by kinetic asymmetries. A lack of prosthetic adaptation to different terrains requires kinematic compensations, which may influence these asymmetries. Six SIGAM grade E-F trans-tibial amputees (one bilateral) wore motion capture markers while standing on force plates, facing down a 5° slope. The participants were tested under three prosthetic conditions; a fixed attachment foot (FIX), a hydraulic ankle (HYD) and a microprocessor foot with a 'standing support' mode (MPF). The resultant ground reaction force (GRF) and support moment for prosthetic and sound limbs were chosen as outcome measures. These were compared between prosthetic conditions and to previously captured able-bodied control data. The distribution of GRF between sound and prosthetic limbs was not significantly affected by foot type. However, the MPF condition required fewer kinematic compensations, leading to a reduction in sound side support moment of 59% (p=0.001) and prosthetic side support moment of 43% (p=0.02) compared to FIX. For the bilateral participant, only the MPF positioned the GRF vector anterior to the knees, reducing the demand on the residual joints to maintain posture. For trans-tibial amputees, loading on lower limb joints is affected by prosthetic foot technology, due to the kinematic compensations required for slope adaptation. MPFs with 'standing support' might be considered reasonable and necessary for bilateral amputees, or amputees with stability problems due to the reduced biomechanical compensations evident.

Quantifying Soft Tissue Artefacts and Imaging Variability in Motion Capture of the Fingers

This study assessed the accuracy of marker-based kinematic analysis of the fingers, considering soft tissue artefacts (STA) and marker imaging uncertainty. We collected CT images of the hand from healthy volunteers with fingers in full extension, mid- and full-flexion, including motion capture markers. Bones and markers were segmented and meshed. The bone meshes for each volunteer’s scans were aligned using the proximal phalanx to study the proximal interphalangeal joint (PIP), and using the middle phalanx to study the distal interphalangeal joint (DIP). The angle changes between positions were extracted. The HAWK protocol was used to calculate PIP and DIP joint flexion angles in each position based on the marker centroids. Finally the marker locations were ‘corrected’ relative to the underlying bones, and the flexion angles recalculated. Static and dynamic marker imaging uncertainty was evaluated using a wand. A strong positive correlation was observed between marker- and CT-based joint angle changes with 0.980 and 0.892 regression slopes for PIP and DIP, respectively, and Root Mean Squared Errors below 4°. Notably for the PIP joint, correlation was worsened by STA correction. The 95% imaging uncertainty interval was < ± 1° for joints, and < ± 0.25 mm for segment lengths. In summary, the HAWK marker set’s accuracy was characterised for finger joint flexion angle changes in a small group of healthy individuals and static poses, and was found to benefit from skin movements during flexion.

20 Optic Flow Stimuli Effects on Standing Balance in Young and Older People with Low and High Fall Risk

Abstract Introduction Optic flow stimuli can destabilise postural control, especially those with increased visual field dependence. This study aimed to examine the effects of optic flow stimuli on postural stability between: 1) younger and older adults; and 2) low-falls-risk and high-falls-risk older adults. Methods 76 participants were grouped into ‘young’ (20-40 years, n=25), and ‘old’ (≥65 years, n=51); the latter was further stratified into ‘old-LFR’ (low fall risk, n=27) and ‘old-HFR’ (high fall risk, n=24). Participants stood on a force platform in a dark room with motion capture markers on the head, torso and lower limb. Optic flow stimuli were projected on a large screen encompassing the field of view as moving white dots for 30 seconds in four different conditions: radial expansion, radial contraction, circular (roll vection) anticlockwise and circular clockwise. Postural control was calculated from center of pressure (COP) and lower limb joint and trunk angles; changes between baseline (static dots) and optic flow conditions were compared between groups. Results Optic flow stimuli led young and older participants to shift their COP in the expected direction for each stimulus condition (forward for expansion, backward for contraction, left for anticlockwise, right for clockwise). For radial expansion, significantly increased anteroposterior sway was found in old-LFR and old-HFR (p≤0.001) but not \young (p≥0.17). Similar results were seen for radial contraction. For circular clockwise, significantly increase mediolateral sway was found in old-HFR (p=0.002) but not young (p=0.912) or old-LFR (p=0.749). Results differed for circular anticlockwise, with greater changes seen in anteroposterior sway compared with mediolateral, particularly in old-HFR. No stimuli by group interaction effects were seen in body kinematics. Conclusions Optic flow stimuli have a more destabilising effect on balance in older than younger adults, particularly those at increased risk of falling.

A model-based motion capture marker location refinement approach using inverse kinematics from dynamic trials.

Marker-based motion capture techniques are commonly used to measure human body kinematics. These techniques require an accurate mapping from physical marker position to model marker position. Traditional methods utilize a manual process to achieve marker positions that result in accurate tracking. In this work, we present an optimization algorithm for model marker placement to minimize marker tracking error during inverse kinematics analysis of dynamic human motion. The algorithm sequentially adjusts model marker locations in 3-D relative to the underlying rigid segment. Inverse kinematics is performed for a dynamic motion capture trial to calculate the tracking error each time a marker position is changed. The increase or decrease of the tracking error determines the search direction and number of increments for each marker coordinate. A final marker placement for the model is reached when the total search interval size for every coordinate falls below a user-defined threshold. Individual marker coordinates can be locked in place to prevent the algorithm from overcorrecting for data artifacts such as soft tissue artifact. This approach was used to refine model marker placements for eight able-bodied subjects performing walking trials at three stride frequencies. Across all subjects and stride frequencies, root mean square (RMS) tracking error decreased by 38.4% and RMS tracking error variance decreased by 53.7% on average. The resulting joint kinematics were in agreement with expected values from the literature. This approach results in realistic kinematics with marker tracking errors well below accepted thresholds while removing variance in the model-building procedure introduced by individual human tendencies.

Modification of knee flexion during walking with use of a real-time personalized avatar

Visual feedback is used in different research areas, including clinical science and neuroscience. In this study, we investigated the influence of the visualization of a real-time personalized avatar on gait parameters, focusing on knee flexion during the swing phase. We also studied the impact of the modification of avatar's knee amplitude on kinematic of the knee of healthy subjects. For this purpose, we used an immersive reality treadmill equipment and developed a 3D avatar, with instantly modifiable parameters for knee flexion and extension (acceleration or deceleration). Fourteen healthy young adults, equipped with motion capture markers, were asked to walk at a self-selected pace on the treadmill. A real-time 3D image of their lower limbs was modelized and projected on the screen ahead of them, as if in a walking motion from left to right. The subjects were instructed to continue walking. When we initiated an increase in the knee flexion of the avatar, we observed a similar increase in the subjects' knee flexion. No significant results were observed when the modification involved a decrease in knee flexion. The results and their significance are discussed using theories encompassing empathy, sympathy and sensory re-calibration. The prospect of using this type of modified avatar for stroke rehabilitation is discussed.

112. Changes in flexion/extension range of motion due to oversizing of cervical total disc arthroplasty

BACKGROUND CONTEXT Total disc arthroplasty (TDA) is a popular treatment for cervical degenerative disc because it preserves motion, decreasing the potential for adjacent level disease. However, oversizing the disc may occur due to limited options in artificial disc height and may diminish this motion preservation. In this study, 1- and 2-level disc arthroplasties with mobile core devices (MCD) at C5-C6 and C4-C5 were compared to oversized devices and 2-level fusions. METHODS Eight cadaveric cervical (C2-T1) spine specimens (average age: 43 ± 12 years) were prepared with cranial and caudal ends embedded in bone cement. Each vertebra was instrumented with optoelectronic markers for motion capture. Each specimen was tested using a six degree-of-freedom kinematic spinal simulator up to 2 Nm in three cycles of flexion/extension. All specimens were tested at four intervention points: intact, ALL resection, 1-level TDA @ C5-C6 with an MCD (Mobi-C, Zimmer Biomet), and 2-level TDA @ C4-C6 with MCDs. After these interventions, the specimens were separated into two equal groups (A and B) with two additional interventions: 2-level TDA with overstuffing or increasing the height of the MCD by 1 mm at C5-C6 (Group A) or C4-C5 (Group B), and 2-level fusion @ C4-C6. All moment and motion capture data were imported into Matlab, and the 3rd test cycle was analyzed. Maximum ranges of motion, normalized as percent of intact, were analyzed for the index (C4-C5 and C5-C6) and adjacent (C3-C4 and C6-C7) levels. Two-way ANOVA with one way repeated-measures was performed with post-hoc Tukey's tests to measure changes in maximum ranges of motion. RESULTS Flexion/Extension, Interventions 1-4: No significant differences were observed between all four interventions at C3-C4 and C4-C5. At C5-C6, significantly greater flexion/extension was measured with ALL resection compared to intact and 1-level TDA interventions. Finally, at C6-C7, ALL resection produced significantly higher range of motion vs 2-level TDA. Flexion/Extension, Group A, Interventions 1,3-6: At C3-C4 and C6-C7, no significant differences were measured between any of the interventions. At C4-C5, intactROM was significantly larger than 2-level TDA, overstuffed TDA @ C5-C6, and 2-level fusion @ C4-C6. In addition, 1-level TDAROM was significantly larger than 2-level TDA and 2-level fusion. Finally, ROM for 2-level TDA and overstuffed TDA @ C5-C6 were both significantly larger than 2-level fusion. At C5-C6, intact, 1-level, and 2-level TDA ROM were significantly larger than the overstuffed TDA @ C5-C6 and the 2-level fusion interventions. Also, the overstuffed TDA @ C5-C6 ROM was significantly larger than the 2-level fusion. Flexion/Extension, Group B, Interventions 1,3-6: At C3-C4 and C6-C7, no significant differences were measured between any of the interventions. At C4-C5, Intact ROM was significantly larger than 2-level TDA, overstuffed TDA @ C4-C5, and 2-level fusion. 1- and 2-level TDA ROM were both significantly larger than overstuffed TDA @ C4-C5 and 2-level fusion. Finally, overstuffed TDA @ C4-C5 ROM was significantly larger than 2-level fusion. At C5-C6, ROM of all interventions was significantly larger than 2-level fusion. CONCLUSIONS Mobile-core disc arthroplasty device maintained range of motion at the index levels, and did not increase motion in adjacent levels in both 1- and 2-level disc arthroplasty implantations. In addition, implanting with oversized devices caused significant decreases in range of motion. FDA DEVICE/DRUG STATUS Unavailable from authors at time of publication.

Analysis and Modeling of Rat Gait Biomechanical Deficits in Response to Volumetric Muscle Loss Injury.

There is currently a substantial volume of research underway to develop more effective approaches for the regeneration of functional muscle tissue as treatment for volumetric muscle loss (VML) injury, but few studies have evaluated the relationship between injury and the biomechanics required for normal function. To address this knowledge gap, the goal of this study was to develop a novel method to quantify the changes in gait of rats with tibialis anterior (TA) VML injuries. This method should be sensitive enough to identify biomechanical and kinematic changes in response to injury as well as during recovery. Control rats and rats with surgically-created VML injuries were affixed with motion capture markers on the bony landmarks of the back and hindlimb and were recorded walking on a treadmill both prior to and post-surgery. Data collected from the motion capture system was exported for post-hoc analysis in OpenSim and Matlab. In vivo force testing indicated that the VML injury was associated with a significant deficit in force generation ability. Analysis of joint kinematics showed significant differences at all three post-surgical timepoints and gait cycle phase shifting, indicating augmented gait biomechanics in response to VML injury. In conclusion, this method identifies and quantifies key differences in the gait biomechanics and joint kinematics of rats with VML injuries and allows for analysis of the response to injury and recovery. The comprehensive nature of this method opens the door for future studies into dynamics and musculoskeletal control of injured gait that can inform the development of regenerative technologies focused on the functional metrics that are most relevant to recovery from VML injury.

Automatic real-time gait event detection in children using deep neural networks.

Annotation of foot-contact and foot-off events is the initial step in post-processing for most quantitative gait analysis workflows. If clean force plate strikes are present, the events can be automatically detected. Otherwise, annotation of gait events is performed manually, since reliable automatic tools are not available. Automatic annotation methods have been proposed for normal gait, but are usually based on heuristics of the coordinates and velocities of motion capture markers placed on the feet. These heuristics do not generalize to pathological gait due to greater variability in kinematics and anatomy of patients, as well as the presence of assistive devices. In this paper, we use a data-driven approach to predict foot-contact and foot-off events from kinematic and marker time series in children with normal and pathological gait. Through analysis of 9092 gait cycle measurements we build a predictive model using Long Short-Term Memory (LSTM) artificial neural networks. The best-performing model identifies foot-contact and foot-off events with an average error of 10 and 13 milliseconds respectively, outperforming popular heuristic-based approaches. We conclude that the accuracy of our approach is sufficient for most clinical and research applications in the pediatric population. Moreover, the LSTM architecture enables real-time predictions, enabling applications for real-time control of active assistive devices, orthoses, or prostheses. We provide the model, usage examples, and the training code in an open-source package.

Methodology on Co-Registration of Mri and Optoelectronic Motion Capture Marker Sets: In-Vivo Wrist Case Study

Methodology on Co-Registration of Mri and Optoelectronic Motion Capture Marker Sets: In-Vivo Wrist Case Study

Determination of number and locations of ultrasound sensors needed to accurately track scapular motion in vivo

ABSTRACTUltrasonic imaging is a safe, established and non-invasive method to visualise the position of tissues within the body. Coupling existing skin mounted motion capture markers to ultrasound sensors could allow for precise measurement of bone positions and orientations, without the loss in accuracy due to relative displacement between skin and bone present in standard motion capture techniques. This is especially important in motions where skin-based markers alone are impractical or inaccurate, as for tracking scapular position when shoulder abduction is greater than 80 degrees. It is currently unclear how many ultrasound-coupled skin-mounted markers would be required to provide accurate scapular tracking, and where they should be placed. In this study, ultrasound-coupled skin marker quantities and positions were optimised using Monte Carlo and simulated annealing methods. Monte Carlo simulation was used to determine the average target registration error in simulated scapular plane abduction motions for groups ranging from 1 to 40 ultrasound-coupled sensors. After the range of ideal sensor quantities was identified, a simulated annealing algorithm was used to identify optimal positions for the sensors. Using these positions, a single skin-mounted, ultrasound-coupled sensor placed over the acromion was found to reduce simulated registration errors by 40.5% compared to regular skin-mounted markers. Groups consisting of three or more sensors reduced simulated registration error at maximum scaption to sub-millimeter levels.

Stride-time variability is related to sensorimotor cortical activation during forward and backward walking

Previous research has used functional near-infrared spectroscopy (fNIRS) to show that motor areas of the cortex are activated more while walking backward compared to walking forward. It is also known that head movement creates motion artifacts in fNIRS data. The aim of this study was to investigate cortical activation during forward and backward walking, while also measuring head movement. We hypothesized that greater activation in motor areas while walking backward would be concurrent with increased head movement. Participants performed forward and backward walking on a treadmill. Participants wore motion capture markers on their head to quantify head movement and pressure sensors on their feet to calculate stride-time. fNIRS was placed over motor areas of the cortex to measure cortical activation. Measurements were compared for forward and backward walking conditions. No significant differences in body movement or head movement were observed between forward and backward walking conditions, suggesting that conditional differences in movement did not influence fNIRS results. Stride-time was significantly shorter during backward walking than during forward walking, but not more variable. There were no differences in activation for motor areas of the cortex when outliers were removed. However, there was a positive correlation between stride-time variability and activation in the primary motor cortex. This positive correlation between motor cortex activation and stride-time variability suggests that forward walking variability may be represented in the primary motor cortex.

Bimodal classification of English allophones employing acoustic speech signal and facial motion capture

A method for automatic transcription of English speech into International Phonetic Alphabet (IPA) system is developed and studied. The principal objective of the study is to evaluate to what extent the visual data related to lip reading can enhance recognition accuracy of the transcription of English consonantal and vocalic allophones. To this end, motion capture markers were placed on the faces of seven speakers to obtain lip tracking data synchronized with the audio signal. 32 markers were used, 20 of which were placed on the speaker's inner lips and 4 on a special cap, which served as the point of reference and stabilized the FMC image while post-processing. Speech samples were simultaneously recorded as a list of approximately 300 words in which all English consonantal and vocalic allophones were represented. Different parameterization strategies were tested and the accuracy of vocalic segments recognition in different experiments was analyzed. The process of multimodal feature extraction is explained, the applied feature selection methods are presented and the obtained results are discussed. Further challenges related to the bi-modal feature extraction process and neural network-based decision systems employment are discussed. [Research sponsored by the Polish National Science Centre, Dec. No. 2015/17/B/ST6/01874.]

Principal Geodesic Analysis in the Space of Discrete Shells

AbstractImportant sources of shape variability, such as articulated motion of body models or soft tissue dynamics, are highly nonlinear and are usually superposed on top of rigid body motion which must be factored out. We propose a novel, nonlinear, rigid body motion invariant Principal Geodesic Analysis (PGA) that allows us to analyse this variability, compress large variations based on statistical shape analysis and fit a model to measurements. For given input shape data sets we show how to compute a low dimensional approximating submanifold on the space of discrete shells, making our approach a hybrid between a physical and statistical model. General discrete shells can be projected onto the submanifold and sparsely represented by a small set of coefficients. We demonstrate two specific applications: model‐constrained mesh editing and reconstruction of a dense animated mesh from sparse motion capture markers using the statistical knowledge as a prior.

Asymptomatic Participants With a Femoroacetabular Deformity Demonstrate Stronger Hip Extensors and Greater Pelvis Mobility During the Deep Squat Task.

Background:Cam-type femoroacetabular impingement (FAI) is a femoral head-neck deformity that causes abnormal contact between the femoral head and acetabular rim, leading to pain. However, some individuals with the deformity do not experience pain and are referred to as having a femoroacetabular deformity (FAD). To date, only a few studies have examined muscle activity in patients with FAI, which were limited to gait, isometric and isokinetic hip flexion, and extension tasks.Purpose:To compare (1) hip muscle strength during isometric contraction and (2) lower limb kinematics and muscle activity of patients with FAI and FAD participants with body mass index–matched healthy controls during a deep squat task.Study Design:Controlled laboratory study.Methods:Three groups of participants were recruited: 16 patients with FAI (14 male, 2 female; mean age, 38.5 ± 8.0 years), 18 participants with FAD (15 male, 3 female; mean age, 32.5 ± 7.1 years), and 18 control participants (16 male, 2 female; mean age, 32.8 ± 7.0 years). Participants were outfitted with electromyography electrodes on 6 muscles and reflective markers for motion capture. The participants completed maximal strength tests and performed 5 deep squat trials. Muscle activity and biomechanical variables were extrapolated and compared between the 3 groups using 1-way analysis of variance.Results:The FAD group was significantly stronger than the FAI and control groups during hip extension, and the FAD group had greater sagittal pelvic range of motion and could squat to a greater depth than the FAI group. The FAI group activated their hip extensors to a greater extent and for a longer period of time compared with the FAD group to achieve the squat task.Conclusion:The stronger hip extensors of the FAD group are associated with greater pelvic range of motion, allowing for greater posterior pelvic tilt, possibly reducing the risk of impingement while performing the squat, and resulting in a greater squat depth compared with those with symptomatic FAI.Clinical Relevance:The increased strength of the hip extensors in the FAD group allowed these participants to achieve greater pelvic mobility and a greater squat depth by preventing the painful impingement position. Improving hip extensor strength and pelvic mobility may affect symptoms for patients with FAI.

The effect of wedge and tibial slope angles on knee contact pressure and kinematics following medial opening-wedge high tibial osteotomy

BackgroundHigh tibial osteotomy is a surgical procedure to treat medial compartment osteoarthritis in varus knees. The reported success rates of the procedure are inconsistent, which may be due to sagittal plane alignment of the osteotomy. The objective of this study was to determine the effect of changing tibial slope, for a range of tibial wedge angles in high tibial osteotomy, on knee joint contact pressure location and kinematics during continuous loaded flexion/extension. MethodsSeven cadaveric knee specimens were cycled through flexion and extension in an Oxford knee-loading rig. The osteotomy on each specimen was adjusted to seven clinically relevant wedge and slope combinations. We used pressure sensors to determine the position of the centre of pressure in each compartment of the tibial plateau and infrared motion capture markers to determine tibiofemoral and patellofemoral kinematics. FindingsIn early knee flexion, a 5° increase in tibial slope shifted the centre of pressure in the medial compartment anteriorly by 4.5mm (P≤0.001), (from the neutral slope/wedge position). Increasing the tibial slope also resulted in the tibia translating anteriorly (P≤0.001). InterpretationChanges to the tibial slope during high tibial osteotomy for all tested wedge angles shifted the centre of pressure in both the medial and lateral compartments substantially and altered knee kinematics. Tibial slope should be controlled during high tibial osteotomy to prevent unwanted changes in tibial plateau contact loads.

Real-time labeling of non-rigid motion capture marker sets

Passive optical motion capture is one of the predominant technologies for capturing high fidelity human motion, and is a workhorse in a large number of areas such as bio-mechanics, film and video games. While most state-of-the-art systems can automatically identify and track markers on the larger parts of the human body, the markers attached to the fingers and face provide unique challenges and usually require extensive manual cleanup. In this work we present a robust online method for identification and tracking of passive motion capture markers attached to non-rigid structures. The method is especially suited for large capture volumes and sparse marker sets. Once trained, our system can automatically initialize and track the markers, and the subject may exit and enter the capture volume at will. By using multiple assignment hypotheses and soft decisions, it can robustly recover from a difficult situation with many simultaneous occlusions and false observations (ghost markers). In three experiments, we evaluate the method for labeling a variety of marker configurations for finger and facial capture. We also compare the results with two of the most widely used motion capture platforms: Motion Analysis Cortex and Vicon Blade. The results show that our method is better at attaining correct marker labels and is especially beneficial for real-time applications.

Development and Validation of a Single Wrist Mounted Inertial Sensor for Biomechanical Performance Analysis of an Elite Netball Shot

The primary objective of the game netball is to score more goals than the opposition. Subsequently, there is impetus on improving shooting accuracy to maximize scoring. Understanding the kinematic factors of shooting that lead to scoring success allows for iterative adjustment and optimization of technique. Three-dimensional retro-reflective motion capture has been used to assess kinematics; however, these systems are expensive, require substantial setup and postprocessing time, and cannot be used in game-such as environments. With modern advancements in wearable microelectromechanical systems based technology, the ability to monitor shooting kinematics in the performance environment has become possible. This article evaluates the efficacy of a single wireless inertial measurement unit (IMU) sensor to monitor shooting kinematics, in terms of forearm angle at ball release. Four elite female shooters shot a total of 30 shots each (totaling 120 shots) from three different distances wearing both reference retroreflective motion capture (Vicon) markers and an IMU. To assess whether wearing the IMU had adverse effects on the kinematic shooting chain, a further ten shots each were taken without wearing the IMU. When contrasted with the gold standard reference of retro-reflective motion capture, the IMU sensor overestimated the forearm angle at ball release established by a mean percentage error of 4.03 1.58. Shots with and without the IMU indicated that the IMU did not biomechanically alter the shooting action. Important advantages of using IMU's over motion-capture solutions include that they are ubiquitous, low cost, require minimal user intervention, and can be used in representative training environments under defensive pressure. This information can enhance the understanding of the athlete's distal segment coordination patterns, providing actionable insights to enable performance to athletes and coaching staff.

Spatial and Temporal Patterns in Sequential Precision Reach Movements.

Introduction: Sequential reach tasks are a common component of manual assembly jobs. These tasks typically involve manipulating a work object or material and reaching to successive target locations...

Dynamic assessment of center of pressure measurements from an instrumented AMTI treadmill with controlled precision

With the increasing use of instrumented force treadmills in biomechanical research, it is imperative that the validity of center of pressure (COP) measurements is established. The study aims were to compare an instrumented treadmill's static-belt COP accuracy to that of a floor-embedded platform, develop a novel method to quantify dynamic-belt COP accuracy with controlled precision and perform an initial investigation of how dynamic COP accuracy changes with weight and velocity. Static COP accuracy was assessed by applying a force while moving a rigid rod in a circular clockwise motion at nine positions of interest on the two treadmill and two ground-embedded force plates. Dynamic COP accuracy was assessed for weights (68.0, 102.1, and 136.1kg), applied through a ball bearing of 2.54cm circumference, with peak treadmill belt speeds of 0.5, 0.75, and 1.0m/s. COP accuracy was assessed relative to motion capture marker trajectories. Statically, treadmill COP error was similar to that of the ground-embedded force plates and that reported for other treadmills. Dynamically, COP error appeared to vary systematically with weight and velocity and in the case of anteroposterior COP error, shear force, although testing with a larger number of weights and velocities is needed to fully define the relationship. This novel method can be used to assess any instrumented treadmill's dynamic COP accuracy with controlled precision.

Magneto-Gyro Wearable Sensor Algorithm for Trunk Sway Estimation During Walking and Running Gait

Trunk sway is a critical gait parameter associated with walking stability and joint loading. This paper introduces a novel algorithm to estimate trunk sway based on magnetometer and gyroscope sensor fusion. An initial experiment was performed to determine the optimal placement of the wearable sensor device on the back and a second experiment was performed to characterize the accuracy of the algorithm. Ten human subjects walked, fast walked, and ran on a treadmill with normal, slightly increased or significantly increased trunk sway. Subjects wore a single magneto-inertial measurement unit (IMU) and a standard set of reflective motion-capture markers on their back. Magneto-gyro algorithm trunk sway estimations were compared with estimations from other common wearable sensor algorithms and errors were determined via comparison with trunk sway measured from motion capture. Overall, the magneto-gyro algorithm was the most accurate (RMSE = 1.7 ± 0.7°) followed by algorithms based on a single three-axis magnetometer (RMSE = 2.5 ± 1.8°), IMU gradient descent (RMSE = 2.9 ± 3.4°), a single three-axis gyroscope (RMSE = 3.2 ± 2.4°), magneto-IMU Kalman filter (RMSE = 8.5 ± 5.5°), and a single accelerometer (RMSE = 16.7 ± 11.2°). Optimal placement of a wearable sensor for estimating trunk sway was along the spine between the $T7$ – $T12$ vertebrae. The presented algorithm based on magnetometer and gyroscope sensor fusion could enable more precise trunk sensing for clinical gait applications outside the laboratory.