Multisite Schizophrenia Classification Based on Brainnetome Atlas by Deep Learning

The paper presents large-scale multisite schizophrenia classification using cutting-edge deep learning tools. In this study, we collected 1275 participants at 8 sites, including 662 schizophrenic patients. We combine the fine-grained Brainnetome Atlas to extract features. And we propose the generalized feature-invariant deep neural network framework to ensure the model generalization in automatic diagnosis schizophrenia. Our model evaluate with 10-fold cross-validation and leave-one-site validation prediction, Average accuracy brings 2.7\% gain in classification accuracy against standard leave-one-site-out validation. Besides, our model gives the promising result on all sites classification and great potentials for computer-aided diagnosis of psychiatric disorders with simple and meaningful biomarkers.