Which data should be tracked in forward-dynamic optimisation to best predict muscle forces in a pathological co-contraction case?

The choice of the cost-function for predicting muscle forces during a movement remains a challenge, especially in patients with neuromuscular disorders. Forward dynamics-based optimisations mainly track joint kinematics or torques, combined with a least-excitation criterion. Tracking marker trajectories and/or electromyography (EMG) has rarely been proposed. Our objective was to determine the best tracking objective-function to accurately predict the upper-limb muscle forces. A musculoskeletal model was created and EMG was simulated to obtain a reference movement – a shoulder abduction. A Gaussian noise (mean = 0; standard deviation = 15%) was added to the simulated EMG. Another noise – corresponding to the actual soft tissue artefacts (STA) of experimental shoulder abduction movements – was added to the trajectories of the markers placed on the model. Muscle forces were estimated from these noisy data, using forward dynamics assisted by six non-linear least-squared objective-functions. These functions involved the tracking of marker trajectories, joint angles or torques, with and without EMG-tracking. All six approaches used the same musculoskeletal model and were solved using a direct multiple shooting algorithm. Finally, the predicted joint angles, muscle forces and activations were compared to the reference values, using root-mean-square errors (RMSe) and biases. The force RMSe of the approach tracking both marker trajectories and EMG (18.45 ± 12.60 N) was almost five times lower than the one of the approach tracking only joint angles (82.37 ± 66.26 N) or torques (85.10 ± 116.40 N). Therefore, using EMG as a complementary tracking-data in forward dynamics seems to be promising for the estimation of muscle forces.     

Rotation sequence is an important factor in shoulder kinematics. Application to the elite players’ flat serves

The aim of this study was to test three different rotation sequences (YXY, ZXY, and XZY) on the shoulder kinematics (rotations of the humerus relative to the thorax) during an original movement such as the tennis flat serve (FS). Nine elite male and female players performed a minimum of five flat serves. An optoelectronic motion analysis system was used to record the movements. Segment kinematics during each FS was reconstructed from the spatial trajectories of the markers according to ISB recommendations. For each rotation sequence, three angles were reported for the shoulder joint, each corresponding to a rotation component around a defined axis. The occurrence of gimbal lock (GL) and angle amplitude coherences were examined. From these three rotation sequences tested, it appears that the XZY sequence was the only decomposition not to suffer from GL. Moreover, the rotation sequence XZY was found to be coherent for all rotation components. Thus, these results show that the best rotation sequence, from both GL and amplitude coherence points of view, is XZY to describe the shoulder kinematics during the tennis serve.     

Do kinematic models reduce the effects of soft tissue artefacts in skin marker-based motion analysis? An in vivo study of knee kinematics

We investigated the effects of including kinematic constraints in the analysis of knee kinematics from skin markers and compared the result to simultaneously recorded trajectories of bone pin markers during gait of six healthy subjects. The constraint equations that were considered for the knee were spherical and revolute joints, which have been frequently used in musculoskeletal modelling. In the models, the joint centres and joint axes of rotations were optimised from the skin marker trajectories over the trial. It was found that the introduction of kinematic constraints did not reduce the error associated with soft tissue artefacts. The inclusion of a revolute joint constraint showed a statistically significant increase in the mean flexion/extension joint angle error and no statistically significant change for the two other mean joint angle errors. The inclusion of a spherical joint showed a statistically significant increase in the mean flexion/extension and abduction/adduction errors. In addition, when a spherical joint was included, a statistically significant increase in the sum of squared differences between measured marker trajectories and the trajectories of the pin markers in the models was seen. From this, it was concluded that both more advanced knee models as well as models of soft tissue artefacts should be developed before accurate knee kinematics can be calculated from skin markers.     

Double calibration: an accurate, reliable and easy-to-use method for 3D scapular motion analysis

The most recent non-invasive methods for the recording of scapular motion are based on an acromion marker (AM) set and a single calibration (SC) of the scapula in a resting position. However, this method fails to accurately measure scapular kinematics above 90° of arm elevation, due to soft tissue artifacts of the skin and muscles covering the acromion. The aim of this study was to evaluate the accuracy, and inter-trial and inter-session repeatability of a double calibration method (DC) in comparison with SC. The SC and DC data were measured with an optoelectronic system during arm flexion and abduction at different angles of elevation (0-180°). They were compared with palpation of the scapula using a scapula locator. DC data was not significantly different from palpation for 5/6 axes of rotation tested (Y, X, and Z in abduction and flexion), where as SC showed significant differences for 5/6 axes. The root mean square errors ranged from 2.96° to 4.48° for DC and from 6° to 9.19° for SC. The inter-trial repeatability was good to excellent for SC and DC. The inter-session repeatability was moderate to excellent for SC and moderate to good for DC. Coupling AM and DC is an easy-to-use method, which yields accurate and reliable measurements of scapular kinematics for the complete range of arm motion. It can be applied to the measurement of shoulder motion in many fields (sports, orthopaedics, and rehabilitation), especially when large ranges of arm motion are required.     

Artefact-reduced kinematics measurement using a geometric finger model with mixture-prior particle filtering

It is challenging to measure the finger's kinematics of underlying bones in vivo. This paper presents a new method of finger kinematics measurement, using a geometric finger model and several markers deliberately stuck on skin surface. Using a multiple-view camera system, the optimal motion parameters of finger model were estimated using the proposed mixture-prior particle filtering. This prior, consisting of model and marker information, avoids generating improper particles for achieving near real-time performance. This method was validated using a planar fluoroscopy system that worked simultaneously with photographic system. Ten male subjects with asymptomatic hands were investigated in experiments. The results showed that the kinematic parameters could be estimated more accurately by the proposed method than by using only markers. There was 20–40% reduction in skin artefacts achieved for finger flexion/extension. Thus, this profile system can be developed as a tool of reliable kinematics measurement with good applicability for hand rehabilitation.     

A Biomechanical Model of the Scapulothoracic Joint to Accurately Capture Scapular Kinematics during Shoulder Movements

The complexity of shoulder mechanics combined with the movement of skin relative to the scapula makes it difficult to measure shoulder kinematics with sufficient accuracy to distinguish between symptomatic and asymptomatic individuals. Multibody skeletal models can improve motion capture accuracy by reducing the space of possible joint movements, and models are used widely to improve measurement of lower limb kinematics. In this study, we developed a rigid-body model of a scapulothoracic joint to describe the kinematics of the scapula relative to the thorax. This model describes scapular kinematics with four degrees of freedom: 1) elevation and 2) abduction of the scapula on an ellipsoidal thoracic surface, 3) upward rotation of the scapula normal to the thoracic surface, and 4) internal rotation of the scapula to lift the medial border of the scapula off the surface of the thorax. The surface dimensions and joint axes can be customized to match an individual’s anthropometry. We compared the model to “gold standard” bone-pin kinematics collected during three shoulder tasks and found modeled scapular kinematics to be accurate to within 2mm root-mean-squared error for individual bone-pin markers across all markers and movement tasks. As an additional test, we added random and systematic noise to the bone-pin marker data and found that the model reduced kinematic variability due to noise by 65% compared to Euler angles computed without the model. Our scapulothoracic joint model can be used for inverse and forward dynamics analyses and to compute joint reaction loads. The computational performance of the scapulothoracic joint model is well suited for real-time applications; it is freely available for use with OpenSim 3.2, and is customizable and usable with other OpenSim models.     

Kalman smoothing improves the estimation of joint kinematics and kinetics in marker-based human gait analysis

We developed a Kalman smoothing algorithm to improve estimates of joint kinematics from measured marker trajectories during motion analysis. Kalman smoothing estimates are based on complete marker trajectories. This is an improvement over other techniques, such as the global optimisation method (GOM), Kalman filtering, and local marker estimation (LME), where the estimate at each time instant is only based on part of the marker trajectories. We applied GOM, Kalman filtering, LME, and Kalman smoothing to marker trajectories from both simulated and experimental gait motion, to estimate the joint kinematics of a ten segment biomechanical model, with 21 degrees of freedom. Three simulated marker trajectories were studied: without errors, with instrumental errors, and with soft tissue artefacts (STA). Two modelling errors were studied: increased thigh length and hip centre dislocation. We calculated estimation errors from the known joint kinematics in the simulation study. Compared with other techniques, Kalman smoothing reduced the estimation errors for the joint positions, by more than 50% for the simulated marker trajectories without errors and with instrumental errors. Compared with GOM, Kalman smoothing reduced the estimation errors for the joint moments by more than 35%. Compared with Kalman filtering and LME, Kalman smoothing reduced the estimation errors for the joint accelerations by at least 50%. Our simulation results show that the use of Kalman smoothing substantially improves the estimates of joint kinematics and kinetics compared with previously proposed techniques (GOM, Kalman filtering, and LME) for both simulated, with and without modelling errors, and experimentally measured gait motion.     

Main component of soft tissue artifact of the upper-limbs with respect to different functional,daily life and sports movements

Soft tissue artifact (STA) is the main source of error in kinematic estimation of human movements based on skin markers. Our objective was to determine the components of marker displacements that best describe STA of the shoulder and arm (i.e. clavicle, scapula and humerus). Four participants performed arm flexion and rotation, a daily-life and a sports movement. Three pins with reflective markers were inserted into the clavicle, scapula and humerus. In addition, up to seven skin markers were stuck on each segment. STA was described with a modal approach: individual marker displacements or marker-cluster (i.e. translations, rotations, homotheties and stretches) relative to the local segment coordinate system defined by markers secured to the pins. The modes were then ranked according to the percentage of total STA energy that they explained. Both individual skin marker displacements and marker-cluster geometrical transformations were task-, location-, segment- and subject-specific. However, 85% of the total STA energy was systematically explained by the rigid transformations (i.e. translations and rotations of the marker-cluster). In conclusion, large joint dislocations and limited efficiency of least squares bone pose estimators are expected for the computation of upper limb joint kinematics from skin markers. Future developments shall consider the rigid transformations of marker-clusters in the implementation of an STA model to reduce its effects on kinematics estimation.     

In situ comparison of A-mode ultrasound tracking system and skin-mounted markers for measuring kinematics of the lower extremity

Skin-mounted marker based motion capture systems are widely used in measuring the movement of human joints. Kinematic measurements associated with skin-mounted markers are subject to soft tissue artifacts (STA), since the markers follow skin movement, thus generating errors when used to represent motions of underlying bone segments. We present a novel ultrasound tracking system that is capable of directly measuring tibial and femoral bone surfaces during dynamic motions, and subsequently measuring six-degree-of-freedom (6-DOF) tibiofemoral kinematics. The aim of this study is to quantitatively compare the accuracy of tibiofemoral kinematics estimated by the ultrasound tracking system and by a conventional skin-mounted marker based motion capture system in a cadaveric experimental scenario. Two typical tibiofemoral joint models (spherical and hinge models) were used to derive relevant kinematic outcomes. Intra-cortical bone pins equipped with optical markers were inserted in the tibial and femoral bones to serve as a reference to provide ground truth kinematics. The ultrasound tracking system resulted in lower kinematic errors than the skin-mounted markers (the ultrasound tracking system: maximum root-mean-square (RMS) error 3.44° for rotations and 4.88 mm for translations, skin-mounted markers with the spherical joint model: 6.32° and 6.26 mm, the hinge model: 6.38° and 6.52 mm). Our proposed ultrasound tracking system has the potential of measuring direct bone kinematics, thereby mitigating the influence and propagation of STA. Consequently, this technique could be considered as an alternative method for measuring 6-DOF tibiofemoral kinematics, which may be adopted in gait analysis and clinical practice.     

Robotics-based synthesis of human motion

The synthesis of human motion is a complex procedure that involves accurate reconstruction of movement sequences, modeling of musculoskeletal kinematics, dynamics and actuation, and characterization of reliable performance criteria. Many of these processes have much in common with the problems found in robotics research. Task-based methods used in robotics may be leveraged to provide novel musculoskeletal modeling methods and physiologically accurate performance predictions. In this paper, we present (i) a new method for the real-time reconstruction of human motion trajectories using direct marker tracking, (ii) a task-driven muscular effort minimization criterion and (iii) new human performance metrics for dynamic characterization of athletic skills. Dynamic motion reconstruction is achieved through the control of a simulated human model to follow the captured marker trajectories in real-time. The operational space control and real-time simulation provide human dynamics at any configuration of the performance. A new criteria of muscular effort minimization has been introduced to analyze human static postures. Extensive motion capture experiments were conducted to validate the new minimization criterion. Finally, new human performance metrics were introduced to study in details an athletic skill. These metrics include the effort expenditure and the feasible set of operational space accelerations during the performance of the skill. The dynamic characterization takes into account skeletal kinematics as well as muscle routing kinematics and force generating capacities. The developments draw upon an advanced musculoskeletal modeling platform and a task-oriented framework for the effective integration of biomechanics and robotics methods.     

Kinematic evaluation of the finger’s interphalangeal joints coupling mechanism—variability,flexion–extension differences,triggers, locking swanneck deformities,anthropometric correlations

The human finger contains tendon/ligament mechanisms essential for proper control. One mechanism couples the movements of the interphalangeal joints when the (unloaded) finger is flexed with active deep flexor. This study’s aim was to accurately determine in a large finger sample the kinematics and variability of the coupled interphalangeal joint motions, for potential clinical and finger model validation applications. The data could also be applied to humanoid robotic hands. Sixty-eight fingers were measured in seventeen hands in nine subjects. Fingers exhibited great joint mobility variability, with passive proximal interphalangeal hyperextension ranging from zero to almost fifty degrees. Increased measurement accuracy was obtained by using marker frames to amplify finger segment motions. Gravitational forces on the marker frames were not found to invalidate measurements. The recorded interphalangeal joint trajectories were highly consistent, demonstrating the underlying coupling mechanism. The increased accuracy and large sample size allowed for evaluation of detailed trajectory variability, systematic differences between flexion and extension trajectories, and three trigger types, distinct from flexor tendon triggers, involving initial flexion deficits in either proximal or distal interphalangeal joint. The experimental methods, data and analysis should advance insight into normal and pathological finger biomechanics (e.g., swanneck deformities), and could help improve clinical differential diagnostics of trigger finger causes. The marker frame measuring method may be useful to quantify interphalangeal joints trajectories in surgical/rehabilitative outcome studies. The data as a whole provide the most comprehensive collection of interphalangeal joint trajectories for clinical reference and model validation known to us to date.     

Feasibility of using a video-based motion analysis system for measuring thumb kinematics

While several different methods have been used to measure hand kinematics, fluoroscopy is generally considered to be the most accurate. Recently, video-based motion analysis has been developed for the measurement of joint kinematics. This method is versatile, easy to use, and can measure motions dynamically. Surface markers are most commonly used in the video-based motion systems. However, whether the surface markers placed on the thumb accurately represent the true kinematics of the underlying bony segment is questionable.In this study, the feasibility of surface markers to represent thumb kinematics was investigated by fluoroscopy. Both the positions of surface markers and bony landmarks were simultaneous recorded and then digitized. The Ra(2) values comparing the angular changes of the thumb interphalangeal, metacarpal and carpometacarpal joints derived using the surface markers or bony landmarks were 0.9986, 0.9730 and 0.9186 in the flexion/extension plane respectively, 0.8837, 0.9697 and 0.8775 in the abduction/adduction plane; and 0.9884, 0.9643 and 0.9431 in the opposition plane. The ranges, mean and standard deviation of the absolute differences between calculated angles of different marker sets were also compared. These data revealed that the similarities of the two different marker techniques throughout the motion cycle were high. The differences between the two methods were also within clinically allowable range of +/-5 degrees. It is concluded that the application of the video-based motion analysis system with surface markers to thumb kinematics is warranted.     

Segmental motion of forefoot and hindfoot as a diagnostic tool

Segmental motions derived from non-invasive motion analysis are being used to investigate the intrinsic functional behavior of the foot and ankle in health and disease. The goal of this research was to examine the ability of a generic segmented model of the foot to capture and differentiate changes in internal skeletal kinematics due to neuromuscular disease and/or trauma. A robotic apparatus that reproduces the kinematics and kinetics of gait in cadaver lower extremities was employed to produce motion under normal and aberrant neuromuscular activation patterns of tibialis posterior and/or tibialis anterior. Stance phase simulations were conducted on 10 donor limbs while recording three-dimensional kinematic trajectories of (1) skin-mounted markers used clinically to construct segmented foot models, and (2) bone-mounted marker clusters to capture actual internal bone motion as the gold standard for comparison. The models constructed from external marker data were able to differentiate the kinematic behaviors elicited by different neuromuscular conditions in a manner similar to that using the bone-derived data. Measurable differences between internal and externally measured kinematics were small, variable and random across the three axes of rotation and neuromuscular conditions, with a tendency toward more differences noted during early and late stance. Albeit slightly different, three-dimensional motion profiles of the hindfoot and forefoot segments correlated well with internal skeletal motion under all neuromuscular conditions, thereby confirming the utility of measuring segmental motions as a valid means of clinical assessment.     

Motion representation of the long fingers: A proposal for the definitions of new anatomical frames

Despite the availability of the International Society of Biomechanics (ISB) recommendations for the orientation of anatomical frames, no consensus exists about motion representations related to finger kinematics. This paper proposes novel anatomical frames for motion representation of the phalangeal segments of the long fingers. A three-dimensional model of a human forefinger was acquired from a non-pathological fresh-frozen hand. Medical imaging was used to collect phalangeal discrete positions. Data processing was performed using a customized software interface (“lhpFusionBox”) to create a specimen-specific model and to reconstruct the discrete motion path. Five examiners virtually palpated two sets of landmarks. These markers were then used to build anatomical frames following two methods: a reference method following ISB recommendations and a newly-developed method based on the mean helical axis (HA). Motion representations were obtained and compared between examiners. Virtual palpation precision was around 1 mm, which is comparable to results from the literature. The comparison of the two methods showed that the helical axis method seemed more reproducible between examiners especially for secondary, or accessory, motions. Computed Root Mean Square distances comparing methods showed that the ISB method displayed a variability 10 times higher than the HA method. The HA method seems to be suitable for finger motion representation using discrete positions from medical imaging. Further investigations are required before being able to use the methodology with continuous tracking of markers set on the subject?s hand.     

Kinematic analysis of a multi-segment foot model for research and clinical applications: a repeatability analysis

An unbiased understanding of foot kinematics has been difficult to achieve due to the complexity of foot structure and motion. We have developed a protocol for evaluation of foot kinematics during barefoot walking based on a multi-segment foot model. Stereophotogrammetry was used to measure retroreflective markers on three segments of the foot plus the tibia. Repeatability was evaluated between-trial, between-day and between-tester using two subjects and two testers. Subtle patterns and ranges of motion between segments of the foot were consistently detected. We found that repeatability between different days or different testers is primarily subject to variability of marker placement more than inter-tester variability or skin movement. Differences between inter-segment angle curves primarily represent a shift in the absolute value of joint angles from one set of trials to another. In the hallux, variability was greater than desired due to vibration of the marker array used. The method permits objective foot measurement in gait analysis using skin-mounted markers. Quantitative and objective characterisation of the kinematics of the foot during activity is an important area of clinical and research evaluation. With this work we hope to have provided a firm basis for a common protocol for in vivo foot study.     

Soft-tissue artefact assessment during step-up using fluoroscopy and skin-mounted markers

When measuring knee kinematics with skin-mounted markers, soft tissue and structures surrounding the knee hide the actual underlying segment kinematics. Soft-tissue artefacts can be reduced when plate-mounted markers or marker trees are used instead of individual unconstrained mounted markers. The purpose of this study was to accurately quantify the soft-tissue artefacts and to compare two marker cluster fixation methods by using fluoroscopy of knee motion after total knee arthroplasty during a step-up task. Ten subjects participated 6 months after their total knee arthroplasty. The patients were randomised into (1) a plate-mounted marker group and (2) a strap-mounted marker group. Fluoroscopic data were collected during a step-up motion. A three-dimensional model fitting technique was used to reconstruct the in vivo 3-D positions of the markers and the implants representing the bones. The measurement errors associated with the thigh were generally larger (maximum translational error: 17mm; maximum rotational error 12 degrees ) than the measurement errors for the lower leg (maximum translational error: 11mm; maximum rotational error 10 degrees ). The strap-mounted group showed significant more translational errors than the plate-mounted group for both the shank (respectively, 3+/-2.2 and 0+/-2.0mm, p = 0.025) and the thigh (2+/-2.0 and 0+/-5.9mm, p = 0.031). The qualitative conclusions based on interpretation of the calculated estimates of effects within the longitudinal mixed-effects modelling evaluation of the data for the two groups (separately) were effectively identical. The soft-tissue artefacts across knee flexion angle could not be distinguished from zero for both groups. For all cases, recorded soft-tissue artefacts were less variable within subjects than between subjects. The large soft-tissue artefacts, when using clustered skin markers, irrespective of the fixation method, question the usefulness of parameters found with external movement registration and clinical interpretation of stair data in small patient groups.     

Kinematics estimation of straddled movements on high bar from a limited number of skin markers using a chain model

To reduce the effects of skin movement artefacts and apparent joint dislocations in the kinematics of whole body movement derived from marker locations, global optimisation procedures with a chain model have been developed. These procedures can also be used to reduce the number of markers when self-occlusions are hard to avoid. This paper assesses the kinematics precision of three marker sets: 16, 11 and 7 markers, for movements on high bar with straddled piked posture. A three-dimensional person-specific chain model was defined with 9 parameters and 12 degrees of freedom and an iterative procedure optimised the gymnast posture for each frame of the three marker sets. The time histories of joint angles obtained from the reduced marker sets were compared with those from the 16 marker set by means of a root mean square difference measure. Occlusions of medial markers fixed on the lower limb occurred when the legs were together and the pelvis markers disappeared primarily during the piked posture. Despite these occlusions, reconstruction was possible with 16, 11 and 7 markers. The time histories of joint angles were similar; the main differences were for the thigh mediolateral rotation and the knee flexion because the knee was close to full extension. When five markers were removed, the average angles difference was about 3 degrees . This difference increased to 9 degrees for the seven marker set. It is concluded that kinematics of sports movement can be reconstructed using a chain model and a global optimisation procedure for a reduced number of markers.     

Accuracy and repeatability of joint angles measured using a single camera markerless motion capture system

Markerless motion capture systems have developed in an effort to evaluate human movement in a natural setting. However, the accuracy and reliability of these systems remain understudied. Therefore, the goals of this study were to quantify the accuracy and repeatability of joint angles using a single camera markerless motion capture system and to compare the markerless system performance with that of a marker-based system. A jig was placed in multiple static postures with marker trajectories collected using a ten camera motion analysis system. Depth and color image data were simultaneously collected from a single Microsoft Kinect camera, which was subsequently used to calculate virtual marker trajectories. A digital inclinometer provided a measure of ground-truth for sagittal and frontal plane joint angles. Joint angles were calculated with marker data from both motion capture systems using successive body-fixed rotations. The sagittal and frontal plane joint angles calculated from the marker-based and markerless system agreed with inclinometer measurements by <0.5°. The systems agreed with each other by <0.5° for sagittal and frontal plane joint angles and <2° for transverse plane rotation. Both systems showed a coefficient of reliability <0.5° for all angles. These results illustrate the feasibility of a single camera markerless motion capture system to accurately measure lower extremity kinematics and provide a first step in using this technology to discern clinically relevant differences in the joint kinematics of patient populations.     

Kinematic models of the upper limb joints for multibody kinematics optimisation: An overview

Soft tissue artefact (STA), i.e. the motion of the skin, fat and muscles gliding on the underlying bone, may lead to a marker position error reaching up to 8.7 cm for the particular case of the scapula. Multibody kinematics optimisation (MKO) is one of the most efficient approaches used to reduce STA. It consists in minimising the distance between the positions of experimental markers on a subject skin and the simulated positions of the same markers embedded on a kinematic model. However, the efficiency of MKO directly relies on the chosen kinematic model. This paper proposes an overview of the different upper limb models available in the literature and a discussion about their applicability to MKO.The advantages of each joint model with respect to its biofidelity to functional anatomy are detailed both for the shoulder and the forearm areas. Models capabilities of personalisation and of adaptation to pathological cases are also discussed. Concerning model efficiency in terms of STA reduction in MKO algorithms, a lack of quantitative assessment in the literature is noted. In priority, future studies should concern the evaluation and quantification of STA reduction depending on upper limb joint constraints.     

In-vivo measurement of dynamic joint motion using high speed biplane radiography and CT: application to canine ACL deficiency

Dynamic assessment of three-dimensional (3D) skeletal kinematics is essential for understanding normal joint function as well as the effects of injury or disease. This paper presents a novel technique for measuring in-vivo skeletal kinematics that combines data collected from high-speed biplane radiography and static computed tomography (CT). The goals of the present study were to demonstrate that highly precise measurements can be obtained during dynamic movement studies employing high frame-rate biplane video-radiography, to develop a method for expressing joint kinematics in an anatomically relevant coordinate system and to demonstrate the application of this technique by calculating canine tibio-femoral kinematics during dynamic motion. The method consists of four components: the generation and acquisition of high frame rate biplane radiographs, identification and 3D tracking of implanted bone markers, CT-based coordinate system determination, and kinematic analysis routines for determining joint motion in anatomically based coordinates. Results from dynamic tracking of markers inserted in a phantom object showed the system bias was insignificant (-0.02 mm). The average precision in tracking implanted markers in-vivo was 0.064 mm for the distance between markers and 0.31 degree for the angles between markers. Across-trial standard deviations for tibio-femoral translations were similar for all three motion directions, averaging 0.14 mm (range 0.08 to 0.20 mm). Variability in tibio-femoral rotations was more dependent on rotation axis, with across-trial standard deviations averaging 1.71 degrees for flexion/extension, 0.90 degree for internal/external rotation, and 0.40 degree for varus/valgus rotation. Advantages of this technique over traditional motion analysis methods include the elimination of skin motion artifacts, improved tracking precision and the ability to present results in a consistent anatomical reference frame.