NeuroFlex
Feasibility of EEG-Based Motor Imagery Control of a Soft Glove for Hand Rehabilitation
Introduction
NeuroFlex is a motion-intent-controlled soft robotic glove developed for hand rehabilitation. Utilizing EEG-based motor imagery (MI) signals and a transformer-based deep learning architecture, NeuroFlex decodes motion intent from EEG data to control a pneumatic glove. This system enables effective rehabilitation through non-invasive brain-computer interaction.
Features
EEG-based Control: Uses MI EEG signals to control the glove for hand movements.
Transformer-Based Model: Implements a self-attention mechanism for accurate EEG signal classification.
Soft Robotic Glove: Lightweight and flexible glove designed with pneumatic actuators for natural movements.
High Accuracy: Achieves up to 85.3% accuracy in classifying MI tasks.