IMU-based real-time Gait Phase Detection algorithm for wearable knee devices

Robotic devices can be used in medical applications for lower-limb impairment such as remobilising paralysed patients, gait retraining and rehabilitation. These devices have limited application outside a clinical setting, and are not suitable for patients with some control over their limbs. Gait phase detection plays a crucial role in the control process to interpret and reinforce the user’s motion.

This research project is aimed to develop a suitable Gait Phase Detection algorithm for real-time application with IMU modules on a knee brace for walking and stair climbing activities. Due to individual gait variations, the algorithms should be adaptive and parameters updated to suit. The project will also develop a separate sensory system to carry out semi-supervised training for detection algorithms.