Dual Joint Visual Feedback Gait Retraining in Pediatric Hemiplegic Cerebral Palsy

Impact and Feasibility of a Dual Joint Visual Feedback Design in Pediatrics With Hemiplegic Cerebral Palsy: A Pilot Study

This study aims to assess the feasibility of a visual feedback system designed to augment treadmill-based gait training for children with gait disturbances related to hemiplegic cerebral palsy. The visual feedback uses data from knee and hip joint movements to represent the position of the foot ahead of or behind the body during walking. This study will test eight children with hemiplegic cerebral palsy who have short step lengths related to atypical knee and hip motions when the foot contacts the ground (initial contact). The study will examine walking adaptations in response to the new visual feedback system and compare differences in response and user experience between two variants of the new design. In a single visit, participants will undergo a gait retraining protocol using the dual joint visual feedback (DJVF) system. Hip and knee flexion angles of the paretic (weaker) side will be collected, analyzed, and compared during baseline walking and while responding to the two feedback variants. Muscle firing patterns will be studied using electrodes placed on key lower limb muscles. Study participants will be asked to report their experience and preferences following the gait retraining protocol. The study findings will be used to further refine the DJVF system in preparation for future studies.

Participant characteristics This study will recruit 8 pediatric participants with hemiplegic gait secondary to CP, in accordance with the inclusion and exclusion criteria. Dual Joint Visual Feedback System (DJVF) The DJVF system uses four inertial measurement unit (IMU) sensors (MTW Devkit, Xsens Technologies BV, Enschede, Netherlands) which are placed on the pelvis, shank, thigh, and foot, to measure three-dimensional segment accelerations and orientations during gait. The software application has been developed by our study team using the MATLAB (Mathworks, Natick, MA) programming language and the application programming interface supplied with the Xsens system. Sensor data is relayed back to a controlling PC for processing and feedback display formulation in real-time. The primary feedback is in the form of a cursor representing the position of the foot ahead of or behind the body. Relative foot position is calculated using the segmental angle and lengths of the paretic thigh and shank. The feedback is designed to reward a more anterior foot placement at the initial contact phase of gait by awarding points according to target screen zones that the cursor enters. The zone reached is highlighted and the instantaneous score is displayed. FBC vs FBCS: addition of cumulative scoring: Two feedback conditions representing two variants will be tested. The first (FBC) is as above, with only instantaneous scoring. The FBCS condition will be the same as the FBC condition with the following addition: game context cues of the cumulative score will be provided as visual numerical count displayed to the right of the main feedback interface. Study Protocol Researchers will place inertial sensors on the pelvis, thigh, shank, and shoe of the paretic limb, using elastic wraps and medical tape. Researchers will place surface EMG sensors on the participant's lower extremity paretic side. If any sensors cannot be placed due to the presence of orthotics or other complicating factors, data for that sensor will not be collected. Researchers will check the signal of each sensor, by having participants perform some easy activities such as flexing their knee, squatting, going on their toes, and raising their toes. Following setup, participants will first walk on the study treadmill at a self-selected comfortable speed while baseline (BAS) gait patterns are recorded. After the baseline trial, practice, and orientation to the feedback, including the FBC and FBCS variants, will occur. The order of these training blocks will be randomized prior to participant arrival. Between training blocks, participants will be allotted rest periods to reduce fatigue. Following testing, participants will complete a questionnaire with items on a five-point ordinal scale describing their 1) comprehension of the feedback display, 2) self-rating of success, 3) preference of which feedback variant (intervention) they preferred, and 4) effort or exhaustion following training.

All participants will take part in the same interventional model. Participants will undergo several gait retraining bouts using two variants of the DJVF system. The order of bouts during the gait retraining will be block randomized.

Participants will undergo a single testing session of gait retraining using two variants of the dual joint visual feedback system (DJVF). The DJVF system uses IMU sensors to measure knee and hip joint motions and calculates the users foot position, relative to their pelvis. The relative foot position is represented in real time on a screen in front of the user with feedback designed to elicit increases in anteriorly located foot positions.

Participants will walk on the treadmill while given instantaneous anterior and posterior foot position and immediate scoring information using the DJVF interface. A pediatric physical therapist will provide (1) verbal encouragement and redirection of attention to the task as needed (2) verbal cueing to assist if difficulty in interpreting the feedback and appropriate modification is noted.

Participants will walk on the treadmill using the same DJFB with additional visual cumulative scoring on their performance. A pediatric physical therapist will provide (1) verbal encouragement and redirection of attention to the task as needed (2) verbal cueing to assist if difficulty in interpreting the feedback and appropriate modification is noted.

Measurement of how bent the knee is when the foot touches the ground at the end of the swing phase without feedback

Measurement of how bent the knee is when the foot touches the ground at the end of the swing phase during exposure to FBC visual feedback

Measurement of how bent the knee is when the foot touches the ground at the end of the swing phase during exposure to FBCS visual feedback

Measurement of how flexed the hip is when the foot touches the ground at the end of the swing phase without feedback

Measurement of how flexed the hip is when the foot touches the ground at the end of the swing phase during exposure to FBC visual feedback

Measurement of how flexed the hip is when the foot touches the ground at the end of the swing phase during exposure to FBCS visual feedback

Participant responses to a five-point ordinal scale detailing their 1) comprehension of the feedback display, 2) self-rating of success, 3) preference of which feedback variant (intervention) they preferred, and 4) effort or exhaustion following training.

Inclusion Criteria: I1: Diagnosis of hemiplegic cerebral palsy I2: Gait pattern in hemiplegic limb: deficit of knee extension during late swing phase/initial contact I3: Age 7 to 17 years I4: Able to walk on a treadmill without assistive devices I5: Able to understand spoken English at the level needed to: understand and follow instructions for equipment setup, testing and task performance answer questions related to effort and preference be able to understand consent document and provide informed consent Exclusion Criteria: E1: Significant injury (such as fracture) in legs or feet that interferes with their ability to walk E2: Known risk factor for stroke or heart attack while exercising E3: Lower extremity orthopaedic surgery less than 12 months prior to study start E4: Other significant health conditions that may interfere with the participant's ability to complete the study or could interfere with interpretation of the study results.