Deep EEG Classifier CNN-LSTM Approach for Mental State Recognition

This paper aims to investigate the use of EEG signal analysis to identify brainwave patterns that can be mapped to cognitive states for human computer interaction purposes. In this work, we propose an EEG signal classification framework based on a convolutional neural network (CNN) coupled with a long short-term memory (LSTM) layer to process multichannel EEG signals. The dataset is a collection of EEG signals corresponding to various cognitive states of the subject, which have been preprocessed and normalized to enhance the feature extraction. The features were also analyzed using Random Forest, SVM, and Decision Trees for comparison purposes. The results showed that the CNN-LSTM model outperformed other conventional classifiers in terms of accuracy. This work shows the feasibility of using deep learning for real-time cognitive state detection which can be useful in applications such as neuro feedback, assistive technology and BCIs