Stability and sensitivity analysis of effort-dependent CCFES for simulated hemiplegic hand assistance after stroke: computational model

Contralaterally-Controlled Functional Electrical Stimulation (CCFES) is an emerging intervention that has evidence for improving therapy for post-stroke hand hemiplegia more than cyclic stimulation (e.g. hand dexterity). Its key feature is that the amount of paretic hand opening assistance is controlled by the degree of non-paretic hand opening, which enables it to assist hand opening during task practice. However, current CCFES techniques do not prevent participants from reducing volitional effort, which is referred to as "slacking" and does not benefit motor relearning. The current study presents a novel effort-dependent CCFES controller that assists only when detecting volitional hand opening. Simulations were used to investigate the effort-dependent controller's stability and sensitivity to system variability using the first-ever computational model of a CCFES trajectory tracking task. System variability was represented by a range of model parameters, including volitional and stimulated hand opening speed, EMG occlusion, stimulation response (M-Wave) phase shifts, and M-Wave filters: Gram-Schmidt, comb, and blanking. System performance was quantified by tracking error and effort estimation accuracy (signal-to-noise ratio, and linear regression r-squared). Results showed that effort-dependent CCFES was highly stable across varying hand opening parameters, EMG-related conditions, and filtering quality -- with \% changes in stability margins being less than 2.2\%. It was also discovered that Gram-Schmidt filter provided significantly better target tracking performance than comb filter and blanking, but it was also significantly more sensitive across parameter variations. Case scenarios were also simulated to demonstrate the stability and performance of effort-dependent CCFES during a tracking task.