Markerless motion capture provides accurate predictions of ground reaction forces across a range of movement tasks.

Measuring or estimating the forces acting on the human body during movement is critical for determining the biomechanical aspects relating to injury, disease and healthy ageing. In this study we examined whether quantifying whole-body motion (segmental accelerations) using a commercial markerless motion capture system could accurately predict three-dimensional ground reaction force during a diverse range of human movements: walking, running, jumping and cutting. We synchronously recorded 3D ground reaction forces (force instrumented treadmill or in-ground plates) with high-resolution video from eight cameras that were spatially calibrated relative to a common coordinate system. We used a commercially available software to reconstruct whole body motion, along with a geometric skeletal model to calculate the acceleration of each segment and hence the whole-body centre of mass and ground reaction force across each movement task. The average root mean square difference (RMSD) across all three dimensions and all tasks was 0.75 N/kg, with the maximum average RMSD being 1.85 N/kg for running vertical force (7.89 % of maximum). There was very strong agreement between peak forces across tasks, with R2 values indicating that the markerless prediction algorithm was able to predict approximately 95-99 % of the variance in peak force across all axes and movements. The results were comparable to previous reports using whole-body marker-based approaches and hence this provides strong proof-of-principle evidence that markerless motion capture can be used to predict ground reaction forces and therefore potentially assess movement kinetics with limited requirements for participant preparation.

Human foot form and function: variable and versatile, yet sufficiently related to predict function from form.

The human foot is a complex structure that plays an important role in our capacity for upright locomotion. Comparisons of our feet with those of our closest extinct and extant relatives have linked shape features (e.g. the longitudinal and transverse arches, heel size and toe length) to specific mechanical functions. However, foot shape varies widely across the human population, so it remains unclear if and how specific shape variants are related to locomotor mechanics. Here we constructed a statistical shape-function model (SFM) from 100 healthy participants to directly explore the relationship between the shape and function of our feet. We also examined if we could predict the joint motion and moments occurring within a person's foot during locomotion based purely on shape features. The SFM revealed that the longitudinal and transverse arches, relative foot proportions and toe shape along with their associated joint mechanics were most variable. However, each of these only accounted for small proportions of the overall variation in shape, deformation and joint mechanics, most likely owing to the high structural complexity of the foot. Nevertheless, a leave-one-out analysis showed that the SFM can accurately predict joint mechanics of a novel foot, based on its shape and deformation.

It is not just the work you do, but how you do it: the metabolic cost of walking uphill and downhill with varying grades.

The cost of walking and running on uneven terrain is not directly explained by external mechanical work. Although metabolic cost of transport increases linearly with gradient at uphill and downhill gradients exceeding 15%, at shallower gradients, the relationship is nonlinear, with the minimum cost occurring at ∼10% downhill grade. Given these nonlinear relationships between grade and metabolic cost, we projected a significant difference in the total metabolic cost of two walking conditions that required the same total external mechanical work be performed over the same total period of time; in one condition, time was spent walking to gradients that were fixed at +10.5% and -10.5% and in the other condition time was spent walking to gradients that varied from 0 to +21% and from -21 to 0%. We compared these two conditions experimentally, using an approach to quantify nonsteady-state oxidative energy expenditure. In line with our projection, the "variable" grade condition resulted in an 8.3 ± 2.2% higher total cumulative oxidative energy expenditure (J·kg-1) compared with the "fixed" grade condition (P < 0.001). Future work should aim to apply our approach across different gradients, speeds, and forms of locomotion; especially those that might provide insight into how humans optimize locomotion on variable grade routes.NEW & NOTEWORTHY We use a method for quantifying nonsteady-state energetics to show that regardless of whether the same total gain and loss in elevation (i.e., same total external mechanical work) is achieved over the same period of time, the total energy expenditure of different graded walking conditions can vary depending on the grades that are walked at and for how long they are walked at.

Validation of a musculoskeletal model for simulating muscle mechanics and energetics during diverse human hopping tasks.

Computational musculoskeletal modelling has emerged as an alternative, less-constrained technique to indirect calorimetry for estimating energy expenditure. However, predictions from modelling tools depend on many assumptions around muscle architecture and function and motor control. Therefore, these tools need to continue to be validated if we are to eventually develop subject-specific simulations that can accurately and reliably model rates of energy consumption for any given task. In this study, we used OpenSim software and experimental motion capture data to simulate muscle activations, muscle fascicle dynamics and whole-body metabolic power across mechanically and energetically disparate hopping tasks, and then evaluated these outputs at a group- and individual-level against experimental electromyography, ultrasound and indirect calorimetry data. Comparing simulated and experimental outcomes, we found weak to strong correlations for peak muscle activations, moderate to strong correlations for absolute fascicle shortening and mean shortening velocity, and strong correlations for gross metabolic power. These correlations tended to be stronger on a group-level rather than individual-level. We encourage the community to use our publicly available dataset from SimTK.org to experiment with different musculoskeletal models, muscle models, metabolic cost models, optimal control policies, modelling tools and algorithms, data filtering etc. with subject-specific simulations being a focal goal.

Mobility of the human foot's medial arch helps enable upright bipedal locomotion.

Developing the ability to habitually walk and run upright on two feet is one of the most significant transformations to have occurred in human evolution. Many musculoskeletal adaptations enabled bipedal locomotion, including dramatic structural changes to the foot and, in particular, the evolution of an elevated medial arch. The foot's arched structure has previously been assumed to play a central role in directly propelling the center of mass forward and upward through leverage about the toes and a spring-like energy recoil. However, it is unclear whether or how the plantarflexion mobility and height of the medial arch support its propulsive lever function. We use high-speed biplanar x-ray measurements of foot bone motion on seven participants while walking and running and compare their motion to a subject-specific model without arch recoil. We show that regardless of intraspecific differences in medial arch height, arch recoil enables a longer contact time and favorable propulsive conditions at the ankle for walking upright on an extended leg. The generally overlooked navicular-medial cuneiform joint is primarily responsible for arch recoil in human arches. The mechanism through which arch recoil enables an upright ankle posture may have helped drive the evolution of the longitudinal arch after our last common ancestor with chimpanzees, who lack arch plantarflexion mobility during push-off. Future morphological investigations of the navicular-medial cuneiform joint will likely provide new interpretations of the fossil record. Our work further suggests that enabling medial arch recoil in footwear and surgical interventions may be critical for maintaining the ankle's natural propulsive ability.

Linking muscle mechanics to the metabolic cost of human hopping.

Many models have been developed to predict metabolic energy expenditure based on biomechanical proxies of muscle function. However, current models may only perform well for select forms of locomotion, not only because the models are rarely rigorously tested across subtle and broad changes in locomotor task but also because previous research has not adequately characterised different forms of locomotion to account for the potential variability in muscle function and thus metabolic energy expenditure. To help to address the latter point, the present study imposed frequency and height constraints to hopping and quantified gross metabolic power as well as the activation requirements of medial gastrocnemius (MG), lateral gastrocnemius (GL), soleus (SOL), tibialis anterior (TA), vastus lateralis (VL), rectus femoris (RF) and biceps femoris (BF), and the work requirements of GL, SOL and VL. Gross metabolic power increased with a decrease in hop frequency and increase in hop height. There was no hop frequency or hop height effect on the mean electromyography (EMG) data of ankle musculature; however, the mean EMG of VL and RF increased with a decrease in hop frequency and that of BF increased with an increase in hop height. With a reduction in hop frequency, GL, SOL and VL fascicle shortening, fascicle shortening velocity and fascicle to MTU shortening ratio increased, whereas with an increase in hop height, only SOL fascicle shortening velocity increased. Therefore, within the constraints that we imposed, decreases in hop frequency and increases in hop height resulted in increases in metabolic power that could be explained by increases in the activation requirements of knee musculature and/or increases in the work requirements of both knee and ankle musculature.

Flexor hallucis brevis motor unit behavior in response to moderate increases in rate of force development.

Background: Studies on motor unit behaviour with varying rates of force development have focussed predominantly on comparisons between slow and ballistic (i.e., very fast) contractions. It remains unclear how motor units respond to less extreme changes in rates of force development. Here, we studied a small intrinsic foot muscle, flexor hallucis brevis (FHB) where the aim was to compare motor unit discharge rates and recruitment thresholds at two rates of force development. We specifically chose to investigate relatively slow to moderate rates of force development, not ballistic, as the chosen rates are more akin to those that presumably occur during daily activity. Methods: We decomposed electromyographic signals to identify motor unit action potentials obtained from indwelling fine-wire electrodes in FHB, from ten male participants. Participants performed isometric ramp-and-hold contractions from relaxed to 50% of a maximal voluntary contraction. This was done for two rates of force development; one with the ramp performed over 5 s (slow condition) and one over 2.5 s (fast condition). Recruitment thresholds and discharge rates were calculated over the ascending limb of the ramp and compared between the two ramp conditions for matched motor units. A repeated measures nested linear mixed model was used to compare these parameters statistically. A linear repeated measures correlation was used to assess any relationship between changes in recruitment threshold and mean discharge rate between the two conditions. Results: A significant increase in the initial discharge rate (i.e., at recruitment) in the fast (mean: 8.6 ±  2.4 Hz) compared to the slow (mean: 7.8 ± 2.3 Hz) condition (P = 0.027), with no changes in recruitment threshold (P = 0.588), mean discharge rate (P = 0.549) or final discharge rate (P = 0.763) was observed. However, we found substantial variability in motor unit responses within and between conditions. A small but significant negative correlation (R2 = 0.33, P = 0.003) was found between the difference in recruitment threshold and the difference in mean discharge rate between the two conditions. Conclusion: These findings suggest that as force increases for contractions with slower force development, increasing the initial discharge rate of recruited motor units produces the increase in rate of force development, without a change in their recruitment thresholds, mean or final discharge rate. However, an important finding was that for only moderate changes in rate of force development, as studied here, not all units respond similarly. This is different from what has been described in the literature for ballistic contractions in other muscle groups, where all motor units respond similarly to the increase in neural drive. Changing the discharge behaviour of a small group of motor units may be sufficient in developing force at the required rate rather than having the discharge behaviour of the entire motor unit pool change equally.

Foot shape is related to load-induced shape deformations, but neither are good predictors of plantar soft tissue stiffness.

Modern human feet are considered unique among primates in their capacity to transmit propulsive forces and re-use elastic energy. Considered central to both these capabilities are their arched configuration and the plantar aponeurosis (PA). However, recent evidence has shown that their interactions are not as simple as proposed by the theoretical and mechanical models that established their significance. Using three-dimensional foot scans and statistical shape and deformation modelling, we show that the shape of the longitudinal and transverse arches varies widely among the healthy adult population, and that the former is subject to load-induced arch flattening, whereas the latter is not. However, longitudinal arch shape and flattening are only one of the various foot shape-deformation relationships. PA stiffness was also found to vary widely. Yet only a small amount of this variability (approx. 10-18%) was explained by variations in foot shape, deformation and their combination. These findings add to the mounting evidence showing that foot mechanics are complex and cannot be accurately represented by simple models. Especially the interactions between longitudinal arch and PA appear to be far less constrained than originally proposed, most likely due to the many degrees of freedom provided by the structural complexity of our feet.

Neuromechanical adaptations of foot function when hopping on a damped surface.

To preserve motion, humans must adopt actuator-like dynamics to replace energy that is dissipated during contact with damped surfaces. Our ankle plantar flexors are credited as the primary source of work generation. Our feet and their intrinsic foot muscles also appear to be an important source of generative work, but their contributions to restoring energy to the body remain unclear. Here, we test the hypothesis that our feet help to replace work dissipated by a damped surface through controlled activation of the intrinsic foot muscles. We used custom-built platforms to provide both elastic and damped surfaces and asked participants to perform a bilateral hopping protocol on each. We recorded foot motion and ground reaction forces, alongside muscle activation, using intramuscular electromyography from flexor digitorum brevis, abductor hallucis, soleus, and tibialis anterior. Hopping in the Damped condition resulted in significantly greater positive work and contact-phase muscle activation compared with the Elastic condition. The foot contributed 25% of the positive work performed about the ankle, highlighting the importance of the foot when humans adapt to different surfaces.NEW & NOTEWORTHY Adaptable foot mechanics play an important role in how we adjust to elastic surfaces. However, natural substrates are rarely perfectly elastic and dissipate energy. Here, we highlight the important role of the foot and intrinsic foot muscles in contributing to replacing dissipated work on damped surfaces and uncover an important energy-saving mechanism that may be exploited by the designers of footwear and other wearable devices.

The feasibility, validity, and reliability of strain measures in the iliotibial band during isolated muscular contractions.

The iliotibial band (ITB) is a unique anatomical structure that transmits forces from two in-series muscles across the lateral knee. Little is known about how force is transmitted, via ITB strain, in response to muscle activation. We have developed a technique to measure the strain through the distal ITB during isolated contractions of the tensor fascia latae (TFL) muscle, using a Kanade-Lucas-Tomasi ultrasound image tracking algorithm. Here we report: 1) the validity of this method to track ITB tissue displacement; 2) the reliability of tracking ITB strain across multiple contractions (intra-probe placement), tracking attempts (intra-operator), data collection sessions (inter-probe placement), and tracking operators (inter-operator); and 3) the feasibility of this approach to assess differences in strain produced during different TFL contraction levels. Our method was valid for tracking ITB displacement and could be used to determine tissue strain due to isolated muscle contraction. Our method was most reliable when a single operator tracked trials without replacing the ultrasound transducer and when averaging across multiple stimulations. Our method was also able to detect changes in ITB strains resulting from differing levels of muscle activation. In the future, this method could be used to assess how factors like posture and ITB region affect the strain found in the distal ITB.

Flexor digitorum brevis utilizes elastic strain energy to contribute to both work generation and energy absorption at the foot.

The central nervous system utilizes tendon compliance of the intrinsic foot muscles to aid the foot's arch spring, storing and returning energy in its tendinous tissues. Recently, the intrinsic foot muscles have been shown to adapt their energetic contributions during a variety of locomotor tasks to fulfil centre of mass work demands. However, the mechanism by which the small intrinsic foot muscles are able to make versatile energetic contributions remains unknown. Therefore, we examined the muscle-tendon dynamics of the flexor digitorum brevis during stepping, jumping and landing tasks to see whether the central nervous system regulates muscle activation magnitude and timing to enable energy storage and return to enhance energetic contributions. In step-ups and jumps, energy was stored in the tendinous tissue during arch compression; during arch recoil, the fascicles shortened at a slower rate than the tendinous tissues while the foot generated energy. In step-downs and landings, the tendinous tissues elongated more and at greater rates than the fascicles during arch compression while the foot absorbed energy. These results indicate that the central nervous system utilizes arch compression to store elastic energy in the tendinous tissues of the intrinsic foot muscles to add or remove mechanical energy when the body accelerates or decelerates. This study provides evidence for an adaptive mechanism to enable the foot's energetic versatility and further indicates the value of tendon compliance in distal lower limb muscle-tendon units in locomotion.

A Human-Centered Machine-Learning Approach for Muscle-Tendon Junction Tracking in Ultrasound Images.

Biomechanical and clinical gait research observes muscles and tendons in limbs to study their functions and behaviour. Therefore, movements of distinct anatomical landmarks, such as muscle-tendon junctions, are frequently measured. We propose a reliable and time efficient machine-learning approach to track these junctions in ultrasound videos and support clinical biomechanists in gait analysis. In order to facilitate this process, a method based on deep-learning was introduced. We gathered an extensive dataset, covering 3 functional movements, 2 muscles, collected on 123 healthy and 38 impaired subjects with 3 different ultrasound systems, and providing a total of 66864 annotated ultrasound images in our network training. Furthermore, we used data collected across independent laboratories and curated by researchers with varying levels of experience. For the evaluation of our method a diverse test-set was selected that is independently verified by four specialists. We show that our model achieves similar performance scores to the four human specialists in identifying the muscle-tendon junction position. Our method provides time-efficient tracking of muscle-tendon junctions, with prediction times of up to 0.078 seconds per frame (approx. 100 times faster than manual labeling). All our codes, trained models and test-set were made publicly available and our model is provided as a free-to-use online service on https://deepmtj.org/.

Mechanics and energetics of human feet: a contemporary perspective for understanding mobility impairments in older adults.

Much of our current understanding of age-related declines in mobility has been aided by decades of investigations on the role of muscle-tendon units spanning major lower extremity joints (e.g., hip, knee and ankle) for powering locomotion. Yet, mechanical contributions from foot structures are often neglected. This is despite the emerging evidence for their critical importance in youthful locomotion. With rapid growth in the field of human foot biomechanics over the last decade, our theoretical knowledge of young asymptomatic feet has transformed, from long-held views of a stiff lever and a shock-absorber to a versatile system that can modulate mechanical power and energy output to accommodate various locomotor task demands. In this perspective review, we predict that the next set of impactful discoveries related to locomotion in older adults will emerge by integrating the novel tools and approaches that are currently transforming the field of human foot biomechanics. By illuminating the functions of feet in older adults, we envision that future investigations will refine our mechanistic understanding of mobility deficits affecting our aging population, which may ultimately inspire targeted interventions to rejuvenate the mechanics and energetics of locomotion.

Stiffening the human foot with a biomimetic exotendon.

Shoes are generally designed protect the feet against repetitive collisions with the ground, often using thick viscoelastic midsoles to add in-series compliance under the human. Recent footwear design developments have shown that this approach may also produce metabolic energy savings. Here we test an alternative approach to modify the foot-ground interface by adding additional stiffness in parallel to the plantar aponeurosis, targeting the windlass mechanism. Stiffening the windlass mechanism by about 9% led to decreases in peak activation of the ankle plantarflexors soleus (~ 5%, p < 0.001) and medial gastrocnemius (~ 4%, p < 0.001), as well as a ~ 6% decrease in positive ankle work (p < 0.001) during fixed-frequency bilateral hopping (2.33 Hz). These results suggest that stiffening the foot may reduce cost in dynamic tasks primarily by reducing the effort required to plantarflex the ankle, since peak activation of the intrinsic foot muscle abductor hallucis was unchanged (p = 0.31). Because the novel exotendon design does not operate via the compression or bending of a bulky midsole, the device is light (55 g) and its profile is low enough that it can be worn within an existing shoe.

The energetic function of the human foot and its muscles during accelerations and decelerations.

The human foot is known to aid propulsion by storing and returning elastic energy during steady-state locomotion. While its function during other tasks is less clear, recent evidence suggests the foot and its intrinsic muscles can also generate or dissipate energy based on the energetic requirements of the center of mass during non-steady-state locomotion. In order to examine contributions of the foot and its muscles to non-steady-state locomotion, we compared the energetics of the foot and ankle joint while jumping and landing before and after the application of a tibial nerve block. Under normal conditions, energetic contributions of the foot rose as work demands increased, while the relative contributions of the foot to center of mass work remained constant with increasing work demands. Under the nerve block, foot contributions to both jumping and landing decreased. Additionally, ankle contributions were also decreased under the influence of the block for both tasks. Our results reinforce findings that foot and ankle function mirror the energetic requirements of the center of mass and provide novel evidence that foot contributions remain relatively constant under increasing energetic demands. Also, while the intrinsic muscles can modulate the energetic capacity of the foot, their removal accounted for only a 3% decrement in total center of mass work. Therefore, the small size of intrinsic muscles appears to limit their capacity to contribute to center of mass work. However, their role in contributing to ankle work capacity is likely important for the energetics of movement.

Neuromechanical adaptations of foot function to changes in surface stiffness during hopping.

Humans choose work-minimizing movement strategies when interacting with compliant surfaces. Our ankles are credited with stiffening our lower limbs and maintaining the excursion of our body's center of mass on a range of surface stiffnesses. We may also be able to stiffen our feet through an active contribution from our plantar intrinsic muscles (PIMs) on such surfaces. However, traditional modeling of the ankle joint has masked this contribution. We compared foot and ankle mechanics and muscle activation on low, medium, and high stiffness surfaces during bilateral hopping using a traditional and anatomical ankle model. The traditional ankle model overestimated work and underestimated stiffness compared with the anatomical model. Hopping on a low stiffness surface resulted in less longitudinal arch compression with respect to the high stiffness surface. However, because midfoot torque was also reduced, midfoot stiffness remained unchanged. We observed lower activation of the PIMs, soleus, and tibialis anterior on the low and medium stiffness conditions, which paralleled the pattern we saw in the work performed by the foot and ankle. Rather than performing unnecessary work, participants altered their landing posture to harness the energy stored by the sprung surface in the low and medium conditions. These findings highlight our preference to minimize mechanical work when transitioning to compliant surfaces and highlight the importance of considering the foot as an active, multiarticular, part of the human leg.NEW & NOTEWORTHY When seeking to understand how humans adapt their movement to changes in substrate, the role of the human foot has been neglected. Using multi-segment foot modeling, we highlight the importance of adaptable foot mechanics in adjusting to surfaces of different compliance. We also show, via electromyography, that the adaptations are under active muscular control.

The extensibility of the plantar fascia influences the windlass mechanism during human running.

The arch of the human foot is unique among hominins as it is compliant at ground contact but sufficiently stiff to enable push-off. These behaviours are partly facilitated by the ligamentous plantar fascia whose role is central to two mechanisms. The ideal windlass mechanism assumes that the plantar fascia has a nearly constant length to directly couple toe dorsiflexion with a change in arch shape. However, the plantar fascia also stretches and then shortens throughout gait as the arch-spring stores and releases elastic energy. We aimed to understand how the extensible plantar fascia could behave as an ideal windlass when it has been shown to strain throughout gait, potentially compromising the one-to-one coupling between toe arc length and arch length. We measured foot bone motion and plantar fascia elongation using high-speed X-ray during running. We discovered that toe plantarflexion delays plantar fascia stretching at foot strike, which probably modifies the distribution of the load through other arch tissues. Through a pure windlass effect in propulsion, a quasi-isometric plantar fascia's shortening is delayed to later in stance. The plantar fascia then shortens concurrently to the windlass mechanism, likely enhancing arch recoil at push-off.

Regulation of foot and ankle quasi-stiffness during human hopping across a range of frequencies.

Human legs operate like springs with adjustable stiffness during locomotion, improving movement economy and versatility. The potential for the foot to contribute to this spring-like mechanism has been established. However, due to previous modelling approaches assuming a rigid-foot segment, it is unknown if ankle and foot quasi-stiffness can be actively regulated, quasi-stiffness being a measure of joint deformation under an applied load. In this study, we sought to determine if midfoot quasi-stiffness was increased in a similar manner as the ankle with increasing load demands, and if these changes were mediated by increased activation of the intrinsic foot muscles. We also quantified differences in measures of ankle joint quasi-stiffness attributable to the modelling assumption of a rigid-foot segment versus a multi-segment foot. Twenty-two healthy individuals performed a single-leg hopping task at three frequencies (2.0 Hz, 2.3 Hz, and 2.6 Hz) while measuring lower limb kinematics, kinetics and muscle activation (EMG). Kinetics and kinematics were calculated at the midfoot and using two different ankle models: rigid-foot model (shank-foot) and an anatomical ankle (shank-calcaneus). Midfoot quasi-stiffness increased with hopping-frequency (p < 0.05), while contact phase intrinsic foot muscle activation decreased (p < 0.05). The assumption of a rigid-foot model overestimated ankle ROM by ~10° and underestimated ankle quasi-stiffness by 45-60%. This study demonstrates that midfoot quasi-stiffness increases with hopping-frequency; however, the mechanism for these changes remains unclear. Furthermore, this study demonstrates the need to assess the ankle and foot independently, using appropriate models, to avoid significant inaccuracies in basic ankle kinematic and kinetic outputs, such as range of motion and joint quasi-stiffness.

The Reliability of Foot and Ankle Bone and Joint Kinematics Measured With Biplanar Videoradiography and Manual Scientific Rotoscoping.

The intricate motion of the small bones of the feet are critical for its diverse function. Accurately measuring the 3-dimensional (3D) motion of these bones has attracted much attention over the years and until recently, was limited to invasive techniques or quantification of functional segments using multi-segment foot models. Biplanar videoradiography and model-based scientific rotoscoping offers an exciting alternative that allows us to focus on the intricate motion of individual bones in the foot. However, scientific rotoscoping, the process of rotating and translating a 3D bone model so that it aligns with the captured x-ray images, is either semi- or completely manual and it is unknown how much human error affects tracking results. Thus, the aim of this study was to quantify the inter- and intra-operator reliability of manually rotoscoping in vivo bone motion of the tibia, talus, and calcaneus during running. Three-dimensional CT bone volumes and high-speed biplanar videoradiography images of the foot were acquired on six participants. The six-degree-of-freedom motions of the tibia, talus, and calcaneus were determined using a manual markerless registration algorithm. Two operators performed the tracking, and additionally, the first operator re-tracked all bones, to test for intra-operator effects. Mean RMS errors were 1.86 mm and 1.90° for intra-operator comparisons and 2.30 mm and 2.60° for inter-operator comparisons across all bones and planes. The moderate to strong similarity values indicate that tracking bones and joint kinematics between sessions and operators is reliable for running. These errors are likely acceptable for defining gross joint angles. However, this magnitude of error may limit the capacity to perform advanced analyses of joint interactions, particularly those that require precise (sub-millimeter) estimates of bone position and orientation. Optimizing the view and image quality of the biplanar videoradiography system as well as the automated tracking algorithms for rotoscoping bones in the foot are required to reduce these errors and the time burden associated with the manual processing.