Detecting Hidden Patterns in EEG Waveforms of Schizophrenia Patients using Convolutional Neural Network

Abstract

Schizophrenia is a severe mental disorder that affects 1% of the world’s population and it is characterized by behavioral symptoms such as delusions, hallucinations and disorganized speech. The aim of this research was to develop an artificial intelligence model to detect hidden patterns in electroencephalogram (EEG) waveforms of schizophrenia patients. EEG waveforms of healthy subjects and schizophrenia patients were collected and processed. The data was used to develop a convolutional neural network (CNN) model which can automatically extract features and classify them. CNN does this by comparing the differences between the EEG waveforms of schizophrenia patients and healthy controls. These differences were used to train the classifier to differentiate the schizophrenia patients from the controls. The result of the CNN model showed a test accuracy of 60%, specificity of 55.55% and a precision of 55.55%. This early result shows that the model is promising. The next step will be to improve the accuracy of the model with a larger pool of data and many iterations, which is expected to lead to a better model that can be relied upon for schizophrenia diagnosis. In conclusion, CNN-based models like this one are relatively cheap and will improve the diagnosis of Schizophrenia, especially in low-income economies where the present study has been carried out.