Abstract
This paper presents an integrated approach to Brain-Computer Interface (BCI) wheelchair development using Artificial Intelligence (AI), leveraging a motor imagery mechanism for controlling right and left-hand movements. The system aims to simulate wheelchair navigation based on motor imagery using electroencephalogram (EEG) data.
A pre-filtered dataset sourced from an open-source EEG repository was segmented into arrays of 19x200 to capture the onset of hand movements, acquired at a sampling frequency of 200Hz. The system integrates a Tkinter-based interface for simulating wheelchair movements, providing users with a functional and intuitive control system.
We propose TFormerEEG, a Transformer-driven deep learning architecture for motor imagery EEG classification. The model achieved a test accuracy of 93.04% when compared to various machine learning baseline models, including XGBoost, EEGNet, and an EEG-Deformer model. The TFormerEEG achieved a mean accuracy of 91.18% through stratified cross-validation, showcasing the effectiveness of this model.
Blogger's Review: This study highlights the potential of Transformer architectures in Brain-Computer Interfaces, particularly in motor imagery tasks. With its high classification accuracy, TFormerEEG not only advances technological progress but also enhances the quality of life for individuals with disabilities, warranting further exploration and optimization.