IMPROVING THE EFFICIENCY OF FES FOR UPPER LIMB REHABILITATION AFTER STROKE

Sarwar-Swarte, Jasmine

Abstract

Research shows high-intensity, functional electrical stimulation (FES)-supported functional task practice is effective for upper limb rehabilitation, promoting neuroplasticity (Rushton, 2003) and improving motor recovery, especially when combined with voluntary effort (Howlett et al., 2015; Sa-e et al., 2020).
FES-UPP is a flexible upper limb (UL) FES system, allowing practice of functional tasks with tailored stimulation levels and voluntary-movement triggered FES (Sun, 2014). However, the FES-UPP system is limited:
-Setup takes approximately 45 minutes (Smith et al., 2019), due, in part, to challenges in manually identifying accurate electrode positions for the desired movement.
-It cannot automatically compensate for fatigue, requiring frequent manual adjustments by therapists (Waring K, 2020).

The PhD aimed to:
-Develop an improved electrode positioning method and
-Develop an adaptive controller which would automatically adjust stimulation levels as the patient fatigued
An experiment was conducted to test a new method for quickly finding optimal electrode positions, with the aim of speeding up UL FES setup times. The proposed method was effective for three out of ten participants for wrist extension, and 5 out of 10 for finger extension. However, position sensitivity varied for all other subjects, so further work is needed.
A fatigue experiment was conducted to characterise FES-related fatigue in healthy participants. Results showed repeatable fatigue rates for the reach forward movement (70 repeats led to a 40% decrease in movement, 100 repeats led to an 80% decrease) in six out of ten subjects, but were inconclusive for all participants for the wrist extension movement.
Lastly, an adaptive control algorithm was developed to reduce time-consuming stimulation adjustments for fatigue compensation while still encouraging voluntary effort. A method to simulate fatigue effects was developed and implemented. The algorithm performed well in pilot testing with one healthy participant.

Key contributions to the field were as follows:
-Simple heuristic methods were proposed to identify the optimal electrode positions which elicit two commonly used, but difficult to isolate, movements. Further work is needed to evaluate these methods in stroke survivors.
-An adaptive controller suitable for use with a state-machine controlled FES system, such as FES-UPP was proposed. The controller performed well in testing with a single healthy subject.

In conclusion, the work has shown the potential to use an adaptive controller to regulate electrical stimulation support over repeated cycles, while still encouraging the patient to attempt to use their own voluntary effort. Future research should assess the algorithm's performance with stroke survivors.

• Thesis Type: Thesis
• Online Publication Date: May 29, 2025
• Deposit Date: May 9, 2025
• Publicly Available Date: May 30, 2027
• Award Date: May 29, 2025