Development of EEG signals acquistion and processing system

Abstract

In this thesis, we present a study on the classification of electroencephalogram (EEG) signals using neural network-based artificial intelligence models to distinguish between healthy individuals and those with mental illnesses. We also created a prototype of an innovative device for collecting EEG signals to assist doctors and researchers in this field. In this work, we focused on a specific and rare disease, which is schizophrenia. Our classification methodology included creating two models, the first using a Convolutional Neural Network (CNN) and the second by adding the Bidirectional Long Short-Term Memory (CNN+Bi-LSTM). The results showed very high potential in the field of classifying complex mental illnesses, as a percentage of 99.39% was recorded using CNN model, outperforming many existing methods and research applied to the same data used.