Motor Intention Recognition from EEG Signals for Augmented Reality-Based Upper-Limb Rehabilitation

Abstract

The goal of this study is to develop a system aimed at supporting the treatment of phantom limb pain by recognizing a user's motor intentions based on brain activity and providing intuitive visual feedback. In particular, a neural network model is designed to classify whether a person is imagining movement of the left or right arm using EEG signals. The detected intention is then visualized through augmented reality (AR) glasses, where virtual arms are displayed and animated according to the imagined movement. Such feedback may help restore the visual perception of limb movement and contribute to reducing phantom pain sensations.