Mental Illness Identification Through EEG Feature Segregation and Machine Learning

This study introduces an advanced machine learning framework for the accurate diagnosis of mental illness such as Schizophrenia (SCZ) using Electroencephalogram (EEG) signals. The research adopts a comprehensive approach that encompasses EEG data acquisition, preprocessing, feature extraction, and classification through convolutional neural networks (CNNs). The experimental dataset consists of 49 SCZ patients and 32 healthy controls. Incorporating feature extraction techniques significantly boosted the diagnostic performance of the model, increasing the accuracy from 68 to 87%. The results indicate that our framework holds considerable promise for the efficient and reliable detection of SCZ, representing a significant advancement in mental health diagnostics.