Control of cyclic movements as muscles fatigue using functional neuromuscular stimulation

In this project, Functional Neuromuscular Stimulation (FNS) was utilized to activate the quadriceps muscle to move the lower leg in a cyclic pattern. During the 300 cycle trials (12.5 minutes), an adaptive neural network, the Pattern Generator/Pattern Shaper (PG/PS), determined the stimulation pattern. Throughout the trials, the adaptation was intermittently mined off and on so that during 3 intervals within the trial, the stimulation pattern was fixed for 50 cycles. We then investigated performance of the controller during intervals with fixed stimulation patterns and intervals with adaptation enabled. Results indicate that the adaptive controller learned to generate an appropriate stimulation pattern during the first interval with adaptation on. The performance of the fixed stimulation pattern then degraded over the next interval, presumably due to muscle fatigue. During subsequent intervals, performance improved with adaptation on and degraded with adaptation off, but the magnitude of the changes were less than during the initial intervals. These results indicate that the adaptive feature was particularly important during initial stages while most of the fatigue was occurring, but that fixed stimulation patterns may be suitable after the fatigue has plateaued.