Invariant hip moment pattern while walking with a robotic hip exoskeleton

Robotic lower limb exoskeletons hold significant potential for gait assistance and rehabilitation; however, we have a limited understanding of how people adapt to walking with robotic devices. The purpose of this study was to test the hypothesis that people reduce net muscle moments about their joints when robotic assistance is provided. This reduction in muscle moment results in a total joint moment (muscle plus exoskeleton) that is the same as the moment without the robotic assistance despite potential differences in joint angles. To test this hypothesis, eight healthy subjects trained with the robotic hip exoskeleton while walking on a force-measuring treadmill. The exoskeleton provided hip flexion assistance from approximately 33% to 53% of the gait cycle. We calculated the root mean squared difference (RMSD) between the average of data from the last 15 min of the powered condition and the unpowered condition. After completing three 30-min training sessions, the hip exoskeleton provided 27% of the total peak hip flexion moment during gait. Despite this substantial contribution from the exoskeleton, subjects walked with a total hip moment pattern (muscle plus exoskeleton) that was almost identical and more similar to the unpowered condition than the hip angle pattern (hip moment RMSD 0.027, angle RMSD 0.134, p<0.001). The angle and moment RMSD were not different for the knee and ankle joints. These findings support the concept that people adopt walking patterns with similar joint moment patterns despite differences in hip joint angles for a given walking speed.     

Paired nonlinear behavior of active and passive joint torques associated with preparation for walk-to-run gait transition

This study investigated the lower extremity torque's active and passive features during the walk-to-run gait transition with continuously increased walking speed. Fourteen volunteers participated in the experiment. Kinematic and kinetic data were collected synchronously. Five strides leading up the gait transition were examined. Peaks of the passive (e.g., contact) and active (e.g., generalized muscle torques), along with net joint torque, and time to peak torques exhibited significant differences at the last stride before gait transition, compared to the first four strides, at the ankle, knee, and hip joints, respectively. Selected peak joint active and passive torques showed significant and opposite trends at critical events within a stride cycle: such ankle joint right after heel-contact, knee joint during weight acceptance, and both hip and knee joints right before toe-off. The magnitude and the corresponding time to active and passive peak torque changed in a nonlinear pattern before the transition from walk to run. The lower extremity segment-interaction during gait transition appeared to be an active reorganization exemplified by the interaction between the lower extremity's active and passive torque components.     

Learning to walk with a robotic ankle exoskeleton

We used a lower limb robotic exoskeleton controlled by the wearer's muscle activity to study human locomotor adaptation to disrupted muscular coordination. Ten healthy subjects walked while wearing a pneumatically powered ankle exoskeleton on one limb that effectively increased plantar flexor strength of the soleus muscle. Soleus electromyography amplitude controlled plantar flexion assistance from the exoskeleton in real time. We hypothesized that subjects' gait kinematics would be initially distorted by the added exoskeleton power, but that subjects would reduce soleus muscle recruitment with practice to return to gait kinematics more similar to normal. We also examined the ability of subjects to recall their adapted motor pattern for exoskeleton walking by testing subjects on two separate sessions, 3 days apart. The mechanical power added by the exoskeleton greatly perturbed ankle joint movements at first, causing subjects to walk with significantly increased plantar flexion during stance. With practice, subjects reduced soleus recruitment by approximately 35% and learned to use the exoskeleton to perform almost exclusively positive work about the ankle. Subjects demonstrated the ability to retain the adapted locomotor pattern between testing sessions as evidenced by similar muscle activity, kinematic and kinetic patterns between the end of the first test day and the beginning of the second. These results demonstrate that robotic exoskeletons controlled by muscle activity could be useful tools for testing neural mechanisms of human locomotor adaptation.     

Effect of body weight support variation on muscle activities during robot assisted gait: a dynamic simulation study

Background and Objectives: While body weight support (BWS) intonation is vital during conventional gait training of neurologically challenged subjects, it is important to evaluate its effect during robot assisted gait training. In the present research we have studied the effect of BWS intonation on muscle activities during robotic gait training using dynamic simulations. Methods: Two dimensional (2-D) musculoskeletal model of human gait was developed conjointly with another 2-D model of a robotic orthosis capable of actuating hip, knee and ankle joints simultaneously. The musculoskeletal model consists of eight major muscle groups namely; soleus (SOL), gastrocnemius (GAS), tibialis anterior (TA), hamstrings (HAM), vasti (VAS), gluteus maximus (GLU), uniarticular hip flexors (iliopsoas, IP), and Rectus Femoris (RF). BWS was provided at levels of 0, 20, 40 and 60% during the simulations. In order to obtain a feasible set of muscle activities during subsequent gait cycles, an inverse dynamics algorithm along with a quadratic minimization algorithm was implemented. Results: The dynamic parameters of the robot assisted human gait such as joint angle trajectories, ground contact force (GCF), human limb joint torques and robot induced torques at different levels of BWS were derived. The patterns of muscle activities at variable BWS were derived and analysed. For most part of the gait cycle (GC) the muscle activation patterns are quite similar for all levels of BWS as is apparent from the mean of muscle activities for the complete GC. Conclusions: Effect of BWS variation during robot assisted gait on muscle activities was studied by developing dynamic simulation. It is expected that the proposed dynamic simulation approach will provide important inferences and information about the muscle function variations consequent upon a change in BWS during robot assisted gait. This information shall be quite important while investigating the influence of BWS intonation on neuromuscular parameters of interest during robotic gait training.     

Biomechanical impairments and gait adaptations post-stroke: Multi-factorial associations

Understanding the potential causes of both reduced gait speed and compensatory frontal plane kinematics during walking in individuals post-stroke may be useful in developing effective rehabilitation strategies. Multiple linear regression analysis was used to select the combination of paretic limb impairments (frontal and sagittal plane hip strength, sagittal plane knee and ankle strength, and multi-joint knee/hip torque coupling) which best estimate gait speed and compensatory pelvic obliquity velocities at toeoff. Compensatory behaviors were defined as deviations from control subjects’ values. The gait speed model (n=18; p=0.003) revealed that greater hip abduction strength and multi-joint coupling of sagittal plane knee and frontal plane hip torques were associated with decreased velocity; however, gait speed was positively associated with paretic hip extension strength. Multi-joint coupling was the most influential predictor of gait speed. The second model (n=15; p<0.001) revealed that multi-joint coupling was associated with increased compensatory pelvic movement at toeoff; while hip extension and flexion and knee flexion strength were associated with reduced frontal plane pelvic compensations. In this case, hip extension strength had the greatest influence on pelvic behavior. The analyses revealed that different yet overlapping sets of single joint strength and multi-joint coupling measures were associated with gait speed and compensatory pelvic behavior during walking post-stroke. These findings provide insight regarding the potential impact of targeted rehabilitation paradigms on improving speed and compensatory kinematics following stroke.     

Persons with multiple sclerosis show altered joint kinetics during walking after participating in elliptical exercise

Patients with multiple sclerosis (MS) experience abnormal gait patterns and reduced physical activity. The purpose of this study was to determine if an elliptical exercise intervention for patients with MS would change joint kinetics during gait toward healthy control values. Gait analysis was performed on patients with MS (n = 24) before and after completion of 15 sessions of supervised exercise. Joint torques and powers were calculated, while also using walking velocity as a covariate, to determine the effects of elliptical exercise on lower extremity joint kinetics during gait. Results show that elliptical exercise significantly altered joint torques at the ankle and hip and joint powers at the ankle during stance. The change in joint power at the ankle indicates that, after training, patients with MS employed a walking strategy that is more similar to that of healthy young adults. These results support the use of elliptical exercise as a gait training tool for patients with MS.     

Biomechanics of overground vs. treadmill walking in healthy individuals.

The goal of this study was to compare treadmill walking with overground walking in healthy subjects with no known gait disorders. Nineteen subjects were tested, where each subject walked on a split-belt instrumented treadmill as well as over a smooth, flat surface. Comparisons between walking conditions were made for temporal gait parameters such as step length and cadence, leg kinematics, joint moments and powers, and muscle activity. Overall, very few differences were found in temporal gait parameters or leg kinematics between treadmill and overground walking. Conversely, sagittal plane joint moments were found to be quite different, where during treadmill walking trials, subjects demonstrated less dorsiflexor moments, less knee extensor moments, and greater hip extensor moments. Joint powers in the sagittal plane were found to be similar at the ankle but quite different at the knee and hip joints. Differences in muscle activity were observed between the two walking modalities, particularly in the tibialis anterior throughout stance, and in the hamstrings, vastus medialis and adductor longus during swing. While differences were observed in muscle activation patterns, joint moments and joint powers between the two walking modalities, the overall patterns in these behaviors were quite similar. From a therapeutic perspective, this suggests that training individuals with neurological injuries on a treadmill appears to be justified.     

The effects of resisted sprint training on acceleration performance and kinematics in soccer, rugby union, and Australian football players

Acceleration is a significant feature of game-deciding situations in the various codes of football. However little is known about the acceleration characteristics of football players, the effects of acceleration training, or the effectiveness of different training modalities. This study examined the effects of resisted sprint (RS) training (weighted sled towing) on acceleration performance (0-15 m), leg power (countermovement jump [CMJ], 5-bound test [5BT], and 50-cm drop jump [50DJ]), gait (foot contact time, stride length, stride frequency, step length, and flight time), and joint (shoulder, elbow, hip, and knee) kinematics in men (N = 30) currently playing soccer, rugby union, or Australian football. Gait and kinematic measurements were derived from the first and second strides of an acceleration effort. Participants were randomly assigned to 1 of 3 treatment conditions: (a) 8-week sprint training of two 1-h sessions x wk(-1) plus RS training (RS group, n = 10), (b) 8-week nonresisted sprint training program of two 1-h sessions x wk(-1) (NRS group, n = 10), or (c) control (n = 10). The results indicated that an 8-week RS training program (a) significantly improves acceleration and leg power (CMJ and 5BT) performance but is no more effective than an 8-week NRS training program, (b) significantly improves reactive strength (50DJ), and (c) has minimal impact on gait and upper- and lower-body kinematics during acceleration performance compared to an 8-week NRS training program. These findings suggest that RS training will not adversely affect acceleration kinematics and gait. Although apparently no more effective than NRS training, this training modality provides an overload stimulus to acceleration mechanics and recruitment of the hip and knee extensors, resulting in greater application of horizontal power.     

The effect of a novel gait retraining device on lower limb kinematics and muscle activation in healthy adults

The Re-Link Trainer (RLT) is a modified walking frame with a linkage system designed to apply a non-individualized kinematic constraint to normalize gait trajectory of the left limb. The premise behind the RLT is that a user’s lower limb is constrained into a physiologically normal gait pattern, ideally generating symmetry across gait cycle parameters and kinematics. This pilot study investigated adaptations in the natural gait pattern of healthy adults when using the RLT compared to normal overground walking. Bilateral lower limb kinematic and electromyography data were collected while participants walked overground at a self-selected speed, followed by walking in the RLT. A series of 2-way analyses of variance examined between-limb and between-condition differences. Peak hip extension and knee flexion were reduced bilaterally when walking in the RLT. Left peak hip extension occurred earlier in the gait cycle when using the RLT, but later for the right limb. Peak hip flexion was significantly increased and occurred earlier for the constrained limb, while peak plantarflexion was significantly reduced. Peak knee flexion and plantarflexion in the right limb occurred later when using the RLT. Significant bilateral reductions in peak electromyography amplitude were evident when walking in the RLT, along with a significant shift in when peak muscle activity was occurring. These findings suggest that the RLT does impose a significant constraint, but generates asymmetries in lower limb kinematics and muscle activity patterns. The large interindividual variation suggests users may utilize differing motor strategies to adapt their gait pattern to the imposed constraint.     

A probabilistic method to estimate gait kinetics in the absence of ground reaction force measurements

Human joint torques during gait are usually computed using inverse dynamics. This method requires a skeletal model, kinematics and measured ground reaction forces and moments (GRFM). Measuring GRFM is however only possible in a controlled environment. This paper introduces a probabilistic method based on probabilistic principal component analysis to estimate the joint torques for healthy gait without measured GRFM. A gait dataset of 23 subjects was obtained containing kinematics, measured GRFM and joint torques from inverse dynamics in order to obtain a probabilistic model. This model was then used to estimate the joint torques of other subjects without measured GRFM. Only kinematics, a skeletal model and timing of gait events are needed. Estimation only takes 0.28 ms per time instant. Using cross-validation, the resulting root mean square estimation errors for the lower-limb joint torques are found to be approximately 0.1 Nm/kg, which is 6–18% of the range of the ground truth joint torques. Estimated joint torque and GRFM errors are up to two times smaller than model-based state-of-the-art methods. Model-free artificial neural networks can achieve lower errors than our method, but are less repeatable, do not contain uncertainty information on the estimates and are difficult to use in situations which are not in the learning set. In contrast, our method performs well in a new situation where the walking speed is higher than in the learning dataset. The method can for example be used to estimate the kinetics during overground walking without force plates, during treadmill walking without (separate) force plates and during ambulatory measurements.     

Muscle contributions to support and progression during single-limb stance in crouch gait

Pathological movement patterns like crouch gait are characterized by abnormal kinematics and muscle activations that alter how muscles support the body weight during walking. Individual muscles are often the target of interventions to improve crouch gait, yet the roles of individual muscles during crouch gait remain unknown. The goal of this study was to examine how muscles contribute to mass center accelerations and joint angular accelerations during single-limb stance in crouch gait, and compare these contributions to unimpaired gait. Subject-specific dynamic simulations were created for ten children who walked in a mild crouch gait and had no previous surgeries. The simulations were analyzed to determine the acceleration of the mass center and angular accelerations of the hip, knee, and ankle generated by individual muscles. The results of this analysis indicate that children walking in crouch gait have less passive skeletal support of body weight and utilize substantially higher muscle forces to walk than unimpaired individuals. Crouch gait relies on the same muscles as unimpaired gait to accelerate the mass center upward, including the soleus, vasti, gastrocnemius, gluteus medius, rectus femoris, and gluteus maximus. However, during crouch gait, these muscles are active throughout single-limb stance, in contrast to the modulation of muscle forces seen during single-limb stance in an unimpaired gait. Subjects walking in crouch gait rely more on proximal muscles, including the gluteus medius and hamstrings, to accelerate the mass center forward during single-limb stance than subjects with an unimpaired gait.     

Invariant ankle moment patterns when walking with and without a robotic ankle exoskeleton

To guide development of robotic lower limb exoskeletons, it is necessary to understand how humans adapt to powered assistance. The purposes of this study were to quantify joint moments while healthy subjects adapted to a robotic ankle exoskeleton and to determine if the period of motor adaptation is dependent on the magnitude of robotic assistance. The pneumatically powered ankle exoskeleton provided plantar flexor torque controlled by the wearer's soleus electromyography (EMG). Eleven naïve individuals completed two 30-min sessions walking on a split-belt instrumented treadmill at 1.25 m/s while wearing the ankle exoskeleton. After two sessions of practice, subjects reduced their soleus EMG activation by ～36% and walked with total ankle moment patterns similar to their unassisted gait (r²=0.98±0.02, THSD, p>0.05). They had substantially different ankle kinematic patterns compared to their unassisted gait (r²=0.79±0.12, THSD, p<0.05). Not all of the subjects reached a steady-state gait pattern within the two sessions, in contrast to a previous study using a weaker robotic ankle exoskeleton (Gordon and Ferris, 2007). Our results strongly suggest that humans aim for similar joint moment patterns when walking with robotic assistance rather than similar kinematic patterns. In addition, greater robotic assistance provided during initial use results in a longer adaptation process than lesser robotic assistance.     

Gait retraining to reduce the knee adduction moment through real-time visual feedback of dynamic knee alignment

Varus knee alignment is a risk factor for medial knee osteoarthritis and is associated with high knee adduction moments. Therefore, reducing the knee adduction moment in varus-aligned individuals with otherwise healthy knees may reduce their risk for developing osteoarthritis. A gait modification that improves dynamic knee alignment may reduce the adduction moment, and systematic training may lead to more natural-feeling and less effortful execution of this pattern. To test these hypotheses, eight healthy, varus-aligned individuals underwent a gait modification protocol. Real-time feedback of dynamic knee alignment was provided over eight training sessions, using a fading paradigm. Natural and modified gait were assessed post-training and after 1 month, and compared to pre-training natural gait. The knee adduction moment, as well as hip adduction, hip internal rotation and knee adduction angles were evaluated. At each training session, subjects rated how effortful and natural-feeling the modified pattern was to execute. Post-training, the modified pattern demonstrated an 8° increase in hip internal rotation and 3° increase in hip adduction. Knee adduction decreased 2°, and the knee adduction moment decreased 19%. Natural gait did not differ between the three visits, nor did the modified gait pattern between the post-training and 1 month visits. The modified pattern felt more natural and required less effort after training. Based on these results, gait retraining to improve dynamic knee alignment resulted in significant reductions in the knee adduction moment, primarily through hip internal rotation. Further, systematic training led to more natural-feeling and less effortful execution of the gait pattern.     

Age causes a redistribution of joint torques and powers during gait.

At self-selected walking speeds, elderly compared with young adults generate decreased joint torques and powers in the lower extremity. These differences may be actual gait-limiting factors and neuromuscular adaptations with age or simply a consciously selected motor pattern to produce a slower gait. The purpose of the study was to compare joint torques and powers of young and elderly adults walking at the same speed. Twelve elderly and fourteen young adults (ages 69 and 21 yr) walked at 1.48 m/s over a force platform while being videotaped. Hip, knee, and ankle torques and powers were calculated from the reaction force and kinematic data. A support torque was calculated as the sum of the three joint torques. Extensor angular impulse during stance and positive work at each joint were derived from the torques and powers. Step length was 4% shorter and cadence was 4% higher in elderly adults (both P < 0.05) compared with young adults. Support angular impulse was nearly identical between groups, but elderly adults had 58% greater angular impulse and 279% more work at the hip, 50% less angular impulse and 39% less work at the knee, and 23% less angular impulse and 29% less work at the ankle compared with young adults (t-test, all P < 0.05). Age caused a redistribution of joint torques and powers, with the elderly using their hip extensors more and their knee extensors and ankle plantar flexors less than young adults when walking at the same speed. Along with a reduction in motor and sensory functions, the natural history of aging causes a shift in the locus of function in motor performance.     

The effects of running in an exerted state on lower extremity kinematics and joint timing

Runners rarely run to the point of maximum fatigue or exhaustion. However, no studies have investigated how the level of exertion associated with a typical running session influences running mechanics. The purpose of this study was to investigate the effects that running in an exerted state had on the kinematics and joint timing within the lower extremity of uninjured, recreational runners. Twenty runners performed a prolonged treadmill run at a self-selected pace that best represented each runner’s typical training run. The run ended based on heart rate or perceived exertion levels that represented a typical training run. Kinematics and joint timing between the foot, knee, and hip were analyzed at the beginning and end of the run. Increases were primarily observed at the end of the run for the peak angles, excursions, and peak velocities of eversion, tibial internal rotation, and knee internal rotation. No differences were observed for knee flexion, hip internal rotation, or any joint timing relationship. Based on these results, runners demonstrated subtle changes in kinematics in the exerted state, most notably for eversion. However, runners were able to maintain joint timing throughout the leg, which may have been a function of the knee. Thus, uninjured runners normally experience small alterations in kinematics when running with typical levels of exertion. It remains unknown how higher levels of exertion influence kinematics with joint timing and the association with running injuries, or how populations with running injuries respond to typical levels of exertion.     

Finger joint coordination during tapping

We investigated finger joint coordination during tapping by characterizing joint kinematics and torques in terms of muscle activation patterns and energy profiles. Six subjects tapped with their index finger on a computer keyswitch as if they were typing on the middle row of a keyboard. Fingertip force, keyswitch position, kinematics of the metacarpophalangeal (MCP) and the proximal and distal interphalangeal (IP) joints, and intramuscular electromyography of intrinsic and extrinsic finger muscles were measured simultaneously. Finger joint torques were calculated based on a closed-form Newton–Euler inverse dynamic model of the finger. During the keystroke, the MCP joint flexed and the IP joints extended before and throughout the loading phase of the contact period, creating a closing reciprocal motion of the finger joints. As the finger lifted, the MCP joint extended and the interphalangeal (IP) joints flexed, creating an opening reciprocal motion. Intrinsic finger muscle and extrinsic flexor activities both began after the initiation of the downward finger movement. The intrinsic finger muscle activity preceded both the IP joint extension and the onset of extrinsic muscle activity. Only extrinsic extensor activity was present as the finger was lifted. While both potential energy and kinetic energy are present and large enough to overcome the work necessary to press the keyswitch, the motor control strategies utilize the muscle forces and joint torques to ensure a successful keystroke.     

Persistence of improvements in postural strategies following motor control training in people with recurrent low back pain.

This study investigated long-term effects of training on postural control using the model of deficits in activation of transversus abdominis (TrA) in people with recurrent low back pain (LBP). Nine volunteers with LBP attended four sessions for assessment and/or training (initial, two weeks, four weeks and six months). Training of repeated isolated voluntary TrA contractions were performed at the initial and two-week session with feedback from real-time ultrasound imaging. Home program involved training twice daily for four weeks. Electromyographic activity (EMG) of trunk and deltoid muscles was recorded with surface and fine-wire electrodes. Rapid arm movement and walking were performed at each session, and immediately after training on the first two sessions. Onset of trunk muscle activation relative to prime mover deltoid during arm movements, and the coefficient of variation (CV) of EMG during averaged gait cycle were calculated. Over four weeks of training, onset of TrA EMG was earlier during arm movements and CV of TrA EMG was reduced (consistent with more sustained EMG activity). Changes were retained at six months follow-up (p<0.05). These results show persistence of motor control changes following training and demonstrate that this training approach leads to motor learning of automatic postural control strategies.     

Optimization-based prediction of asymmetric human gait

An optimization-based formulation and solution method are presented to predict asymmetric human gait for a large-scale skeletal model. Predictive dynamics approach is used in which both the joint angles and joint torques are treated as unknowns in the equations of motion. For the optimization formulation, the joint angle profiles are treated as the primary unknowns, and velocities and accelerations are calculated using them. In numerical implementation, the joint angle profiles are discretized using the B-spline interpolation. An algorithm is presented to inversely calculate the joint torques and the ground reaction forces. The sum of the joint-torques squared, called the dynamic effort, is minimized as the human performance measure. Constraints are imposed on the joint strengths (torques) and joint ranges of motion along with other physical constraints. The formulation is validated by simulating a symmetric gait and comparing the results with the experimental data. Then asymmetric gait motion is simulated, where the left and right step lengths are different. The kinematics and kinetics results from the simulation are presented and discussed. Predicted ground reaction forces are explained by using the inverted pendulum model. Predicted kinematics and kinetics have trends that are similar to those reported in the literature. Potential practical applications of the formulation and the solution approach are discussed.     

Modulation of shoulder muscle and joint function using a powered upper-limb exoskeleton

Robotic-assistive exoskeletons can enable frequent repetitive movements without the presence of a full-time therapist; however, human-machine interaction and the capacity of powered exoskeletons to attenuate shoulder muscle and joint loading is poorly understood. This study aimed to quantify shoulder muscle and joint force during assisted activities of daily living using a powered robotic upper limb exoskeleton (ArmeoPower, Hocoma). Six healthy male subjects performed abduction, flexion, horizontal flexion, reaching and nose touching activities. These tasks were repeated under two conditions: (i) the exoskeleton compensating only for its own weight, and (ii) the exoskeleton providing full upper limb gravity compensation (i.e., weightlessness). Muscle EMG, joint kinematics and joint torques were simultaneously recorded, and shoulder muscle and joint forces calculated using personalized musculoskeletal models of each subject’s upper limb. The exoskeleton reduced peak joint torques, muscle forces and joint loading by up to 74.8% (0.113 Nm/kg), 88.8% (5.8%BW) and 68.4% (75.6%BW), respectively, with the degree of load attenuation strongly task dependent. The peak compressive, anterior and superior glenohumeral joint force during assisted nose touching was 36.4% (24.6%BW), 72.4% (13.1%BW) and 85.0% (17.2%BW) lower than that during unassisted nose touching, respectively. The present study showed that upper limb weight compensation using an assistive exoskeleton may increase glenohumeral joint stability, since deltoid muscle force, which is the primary contributor to superior glenohumeral joint shear, is attenuated; however, prominent exoskeleton interaction moments are required to position and control the upper limb in space, even under full gravity compensation conditions. The modeling framework and results may be useful in planning targeted upper limb robotic rehabilitation tasks.     

Manipulating post-stroke gait: Exploiting aberrant kinematics

Post-stroke individuals often exhibit abnormal kinematics, including increased pelvic obliquity and hip abduction coupled with reduced knee flexion. Prior examinations suggest these behaviors are expressions of abnormal cross-planar coupling of muscle activity. However, few studies have detailed the impact of gait-retraining paradigms on three-dimensional joint kinematics. In this study, a cross-tilt walking surface was examined as a novel gait-retraining construct. We hypothesized that relative to baseline walking kinematics, exposure to cross-tilt would generate significant changes in subsequent flat-walking joint kinematics during affected limb swing. Twelve post-stroke participants walked on a motorized treadmill platform during a flat-walking condition and during a 10-degree cross-tilt with affected limb up-slope, increasing toe clearance demand. Individuals completed 15 min of cross-tilt walking with intermittent flat-walking catch trials and a final washout period (5 min). For flat-walking conditions, we examined changes in pelvic obliquity, hip abduction/adduction and knee flexion kinematics at the spatiotemporal events of swing initiation and toe-off, and the kinematic event of maximum angle during swing. Pelvic obliquity significantly reduced at swing initiation and maximum obliquity in the final catch trial and late washout. Knee flexion significantly increased at swing initiation, toe-off, and maximum flexion across catch trials and late washout. Hip abduction/adduction was not significantly influenced following cross-tilt walking. Significant decrease in the rectus femoris and medial hamstrings muscle activity across catch trials and late washout was observed. Exploiting the abnormal features of post-stroke gait during retraining yielded desirable changes in muscular and kinematic patterns post-training.