Automatic recognition of schizophrenia from brain-network features using graph convolutional neural network

• Published in: Asian journal of psychiatry Vol. 87; p. 103687
• Main Authors: Yin, Guimei, Chang, Ying, Zhao, Yanli, Liu, Chenxu, Yin, Mengzhen, Fu, Yongcan, Shi, Dongli, Wang, Lin, Jin, Lizhong, Huang, Jie, Li, Dandan, Niu, Yan, Wang, Bin, Tan, Shuping
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.09.2023
• Subjects: Epoch length; Frequency band; Functional connectivity metrics; Graph convolutional neural network; Schizophrenia; Schizophrenia; Functional connectivity metrics; Epoch length; Frequency band; Graph convolutional neural network
• ISSN: 1876-2018; 1876-2026
• DOI: 10.1016/j.ajp.2023.103687

Schizophrenia is a severe mental illness that imposes considerable economic burden on families and society. However, its clinical diagnosis primarily relies on scales and doctors' clinical experience and lacks an objective and accurate diagnostic approach. In recent years, graph convolutional neural networks (GCN) have been used to assist in psychiatric diagnosis owing to their ability to learn spatial-association information. Therefore, this study proposes a schizophrenia automatic recognition model based on graph convolutional neural network. Herein, the resting-state electroencephalography (EEG) data of 103 first-episode schizophrenia patients and 92 normal controls (NCs) were obtained. The automatic recognition model was trained with a nodal feature matrix that comprised the time and frequency-domain features of the EEG signals and local features of the brain network. The most significant regions that contributed to the model classification were identified, and the correlation between the node topological features of each significant region and clinical evaluation metrics was explored. Experiments were conducted to evaluate the performance of the model using 10-fold cross-validation. The best performance in the theta frequency band with a 6 s epoch length and phase-locked value. The recognition accuracy was 90.01%. The most significant region for identifying with first-episode schizophrenia patients and NCs was located in the parietal lobe. The results of this study verify the applicability of the proposed novel method for the identification and diagnosis of schizophrenia. •An EFC-GCN model was constructed, and the optimal epoch length, frequency band and functional connection metrics for recognition are derived through feature extraction and classification recognition by GCN, which fully learns the association between EEG channels.•The time and frequency domain features of the EEG signal and the local features of the brain network are extracted to form the node feature matrix, and the information of the time, frequency and topological features of the EEG signal is fully learned.•This study innovatively combines the Grad-CAM algorithm with the EFC-GCN model to identify the most significant brain regions that contribute to the classification, and to explore the correlation between the network topological features of the significant brain regions and the clinical score values.