Advancing Through the Integration of Deep Neural Networks and EEG Analysis: A Data-Driven Approach

As a neurological illness, schizophrenia(sz) may impair brain function and lead to irrational behaviour. Better patient outcomes are achieved by early identification and treatment. Analysis of electroencephalogram (EEG) data has been enhanced by recent improvements in machine learning (ML) methods for feature extraction and categorization. A large body of research has investigated the possibility of classifying EEG data using different machine learning methods. In this research, we analyses the effectiveness of 1D convolutional neural network (1D CNN) model. This research utilized electroencephalogram (EEG) dataset with 28 participants - from 14 schizophrenia (SZ) participant and 14 individuals who were healthy. To improve the data quality, data preparation procedures were used, including filtering and normalisation. The model demonstrates strong performance with a 96% accuracy rate, supported by high sensitivity (97%) and specificity (96%), indicating a well-balanced approach that effectively predicts both classes without bias. The proposed model could serve as an effective diagnostic method for healthcare practitioners in the early identification of SZ.