The research team has developed a visual kinematic biofeedback system which is designed to help children with hemiplegic cerebral palsy (CP) correct a pattern of reduced knee extension in terminal swing and early stance. The system provides real-time feedback on the knee angle pattern during walking on a treadmill. From a pilot study on children with CP, the investigators observed that when the system was used in children who have stiff knee gait (SKG), training with knee feedback alone could lead to an increase in hip flexion which in turn led to limited normalization of the knee pattern through the whole gait cycle. This study, funded by the NIDILRR Switzer grant (PI: X Liu, Ph.D.), seeks to address the question of whether a training design with feedback on both the knee and hip joints would reduce this tendency to generate unintended changes in hip joint motion, and in doing so also improve convergence to the intended knee joint pattern. This study will test ten children and young adults with brain injury who have SKG and examine their short term adaptations to two types of kinematic feedback training: feedback training on the knee alone (condition B) and sequential switched feedback training on the knee and the hip (condition A). An additional sensor placed on the pelvis will be added to the current feedback system for measurement and feedback on the hip joint angle. Software enhancements will also be made with methods that will allow study and description of adaptations in measures of inter-limb symmetry during training. The participants will visit twice with a 2-week washout period between the two visits. Five participants will first undergo condition B in the first visit and then condition A in the second visit, while the other five participants will start with condition A in the first visit and then undergo condition B in the second visit. To compare the effects of the conditions on normalizing the joint angle trajectories, the knee and hip kinematics will be collected and analyzed in both the conditions. To investigate the coordination of lower limb segments under feedback training, relative phase measures will be analyzed on the hip and the knee. To examine whether participants adapt to the feedback retraining in terms of improvement in gait quality, symmetry ratios will be analyzed.

Participant characteristics

This study will recruit 10 participants according to the inclusion and exclusion criteria.

System development

In order to measure the quantitative change in hip joint angle online, an additional sensor will be placed on the pelvis segment (overlying the sacrum between the posterior superior iliac spines) and added to the current feedback system. Totally four sensors will be used, including the sensors on the pelvis, thigh, shank, and heel. The hip flexion angle will be calculated from pelvis and thigh sensors. The knee flexion angle will be calculated from thigh and shank sensors, while the heel sensor signal is monitored to isolate strides by detecting contact of the foot with the support surface. To help subjects easily recognize which joint the feedback is cueing for, different backgrounds is selected for the feedback interface for the hip joint and knee joint, respectively. Gait patterns will be video recorded in a sagittal view of the lower extremities. In order to test gait asymmetry by symmetry ratio (dividing the smaller value by the larger value between trained and untrained lower limbs), stance phase duration (% gait cycle) from heel strike to toe off will be identified by motion capturing system with reflective markers. Reflective markers will be placed on the ankles and shoes (fifth metatarsal, heel, rearfoot along the line from heel to toe and below the ankle) on both sides to measure the heel down and toe off events.

Biofeedback gait retraining protocol

Feedback training on the knee alone (Condition B) will include four 6-mins training blocks: 4-mins knee joint feedback-on and 2-mins feedback-off. Sequential switched feedback training on the knee and the hip (Condition A) will include four 6-mins training blocks: 2-mins knee joint feedback-on, 2-mins hip joint feedback-on, and 2-mins feedback-off. Subjects will wear their comfortable footwear and daily used bracing/orthotic devices during the training sessions. Totally four inertial sensors (MTw, Xsens, Netherlands) will be placed separately on the pelvis, anterior thigh, posterior shank and the heel of the paretic lower limb. For calibration purpose, the subject will first stand in his/her natural standing posture and then stand with the knee (weaker side) flexed at 60 degree adding to the knee angle in the natural standing posture. Prior to training, a treadmill walking trial with comfortable speed will be recorded as the baseline trial. After the baseline trial, a 3 minute practice trial will allow interactive demonstration and practice with the goal that the subject has a clear understanding of the task and how to interpret the feedback. In both the two training conditions, subjects will have 3 mins sitting rest between two consecutive blocks to reduce fatigue.

Inclusion Criteria:

• age 7 to 21;
• has SKG
• diagnosed with brain injury including but not limited to Cerebral Palsy, Stroke,Traumatic Brain Injury;
• ability to walk on a treadmill without assistive devices based on parent/guardian report and/or treatment history;
• the cognitive development is at the level needed to: understand and follow instructions, answer questions, be able to understand the purpose of the study and the activities involved.

Exclusion Criteria:

• Botulinum toxin treatment less than 16 weeks before initiation of the study
• Recent or concurrent treatment that might interfere with the study.