The question
People with tetraplegia may retain individual patterns of shoulder or upper-arm movement and muscle activation. How can these abilities support reliable control of an assistive device or neuroprosthesis?
The approach
My work explores wearable inertial sensing and electromyography to recognise these signals and map them to an action. Personal calibration matters: useful movements and comfortable control strategies differ between individuals.
In the 2022 residual-movement study, a single inertial sensor captured shoulder movement. A support vector machine classified the signal into user commands, with a short user-specific setup and training process.
What the research shows
The published work demonstrates the feasibility of controlling robotic and stimulation-assisted grasp through residual movement or muscle contraction. These are research demonstrations; they do not establish that one interface will be suitable for every person or setting.
The 2022 interface study included ten participants without spinal cord injury and two participants with spinal cord injury. Larger and longer evaluations remain important when considering wider use.
Where this leads
Questions include how interfaces adapt over time, how feedback helps a person learn the control strategy, and how to evaluate reliability alongside effort and agency.