EEG-Based Classification of Schizophrenia and Bipolar Disorder with the Fuzzy Method

Abstract

This study demonstrates various fuzzy-based strategies for classifying and diagnosing people with mental illnesses such as schizophrenia and bipolar disorder. The signals collected from 32 unipolar electrodes during non-invasive electroencephalogram analysis were examined to determine their key characteristics. This research uses a sophisticated fuzzy-based radial basis function neural network. Entropy analysis and analysis of variance of other statistical parameters are also used. Three hundred and twelve schizophrenic patients and 105 individuals with bipolar disorder were examined. In contrast to healthy controls, the data indicated that the patients were correctly classified. With close to 96% accuracy, the suggested method outperforms existing machine learning methods, such as support vector machines and k-nearest neighbors. Conclusion: This categorization method will enable the development of highly accurate algorithms to identify and classify various mental illnesses.