Design And Implementation of a Cost-Effective Hand Exoskeleton Controlled by Brain Computer Interface Using EEG And EMG

This paper presents the development of a cost-effective and portable hand exoskeleton device that utilizes a hybrid Brain-Computer Interface (BCI) system to assist individuals with paralysis in regaining hand movement. Neuro logical impairments often lead to significant motor function restrictions, with millions of people worldwide unable to access advanced rehabilitation devices due to their high costs. The proposed device interprets neural signals to detect user intentions, converting these signals into real-time hand movements, thus bypassing the need for physical muscle activation. A hybrid machine learning model combining Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNN) enhances signal classification accuracy. Due to constraints in acquiring an EEG sensor, the final implementation may integrate either EEG or Electromyography (EMG) sensors, significantly impacting system properties. Designed for accessibility and ease of operation, this device represents a step forward in affordable rehabilitation technology. Preliminary findings suggest high accuracy in intention detection, indicating strong potential for improving motor functionality and independence. This hybrid-driven hand exoskeleton highlights a promising direction in accessible rehabilitation, enabling practical support for everyday tasks and fostering an enhanced quality of life.