Predicting Schizophrenia Using EEG Signals and Convolutional Long Short-Term Memory

Schizophrenia is a mental illness that affects the cognitive and social functions of many individuals and is characterized by impairments in thoughts and social interaction. Treatment is most effective when it occurs early, particularly with an early diagnosis. In this paper, we present a novel deep learning method for predicting schizophrenia using EEG signals and advanced deep learning techniques, such as Convolutional Long Short-Term Memory (ConvLSTM). By leveraging the benefits of convolutional layers and the Long Short-Term Memory (LSTM) network, the system can effectively identify patterns in the EEG signals associated with schizophrenia. The model is trained and evaluated on publicly available EEG datasets, demonstrating excellent capability in classifying schizophrenia and achieving high accuracy. These results show that conventional machine learning models, which are not based on ConvLSTM, fail to provide better performance, whereas ConvLSTM-based models exhibit superior predictive performance.