Abstract of Dr. Helen Huang's Talk

Talk title: Harnessing Reinforcement Learning and Physiological Models for EMG-based Neural-Machine Interface Design

Abstract: 

In this talk, I will present my lab’s research on developing novel EMG-based neural–machine interfaces that integrate musculoskeletal and arm dynamics models (physical and physiological models) within a reinforcement learning framework to estimate coordinated, continuous hand and wrist movements in individuals with intact limbs as well as those with limb loss. Our approach not only enables personalized neural decoding for external assistive device control but also provides a principled way to interpret physical deficits through the parameters and internal

states of the underlying physical and physiological models.