Safety Improvement in Using Automatic Control of FES-assisted Locomotion

This study has focused on the following two issues: 1) development of a method for safety evaluation of FESassisted locomotion for persons with spinal cord injury, and 2) implementation of this method to safety evaluation of the locomotion of persons with incomplete spinal cord injury using automatic control system for FES based on adaptive logic networks (ALNs) and restriction rules (RRs). Discrimination of gait events is required for automatic control of FES used to assist with locomotion after lesions in the central nervous system. We have previously demonstrated that ALNs can be used in gait event discrimination (GED) and automatic design of control algorithms for FES-assisted locomotion [Kostov et al., 1995]. Machine learning techniques for supervised learning generalize well in cloning the skill of the user controlling the stimulation manually. However, safe operation of a GED-based control system applied to a highly variable human walking may require more redundancy than ALNs alone can provide. A practical method was designed to evaluate functional performance of the man-machine system. Functional errors of the FES-control system were defined, and restriction rules (RRs), based on a priori knowledge about the user’s walking, were introduced to verify decisions made by ALNs and to eliminate infrequent functional errors, providing additional safety for the user. Our results demonstrated that RRs can eliminate most of the critical functional errors if a certain level of ALN training is achieved. Our future efforts are directed toward expanding the training error assessment method to define the training level required for error-free performance.