Incorporating Motor Imagery-Controlled Gaming into Paralysis Rehabilitation

The paper presents a realtime neuro-gaming system that uses neural signals to help play games. This gaming system will significantly help in recovering from quadriplegia, which can act as physical therapy for rehabilitation. By simply imagining physical movements or actions, they can interact with computers and get visual feedback on the action through the video game. This can encourage the users to engage in mental exercises that stimulate brain activity and neuroplasticity. The design utilizes advanced deep-learning algorithms to analyze different types of brain signals. The Electroencephalography (EEG) data is structured as a time series, and Long Short-Term Memory (LSTM) has proven efficient in analyzing and classifying this data into three distinct classes. EEG signals have been preprocessed and classified using a Long-short term memory algorithm. The classifier's output has been integrated into the gaming input system, complementing the traditional gaming input system. LSTM has provided satisfactory results in classifying preprocessed realtime EEG data with 72% accuracy. The system processes data and completes classification in realtime. The system is engineered to enable precise signal classification with minimal preprocessing requirements.