Towards Efficient Personalization of Computerized Lower Limb Prostheses Via Reinforcement Learning in a Clinical Setup - Patient Study

What this Trial Is About

Researchers are evaluating the use of reinforcement learning to personalize robotic prosthetic legs for adults with unilateral transfemoral amputation. This clinical trial aims to determine if the RL-based Recommendation Interfacing System (RISE) allows clinicians to personalize prosthetic legs faster and without detailed control knowledge compared to manual methods. The study compares patient performance and personalization speed across different personalization approaches. Participants will experience three personalization methods: tuning experts manually adjusting prosthetic controls, tuning experts using the RISE system to guide adjustments, and prosthetists without tuning expertise using RISE. The study involves standard prosthetic fitting procedures like alignment adjustment, followed by multiple personalization sessions on different terrains. Participants will complete a testing trial where walking performance with each personalization approach is measured. Participants are expected to visit the testing site about eight times: one visit for alignment, three personalization procedures repeated twice each, and one or two visits for the final testing trial. Researchers will record outcomes such as total personalization time during visits, walking speed, step length, and foot elevation during evaluation visits. Safety and adherence will be monitored throughout the study period ending in November 2029.

Towards Efficient Personalization of Computerized Lower Limb Prostheses Via Reinforcement Learning in a Clinical Setup - Group 1