Recent advances of deep learning in psychiatric disorders

Background:A number of brain research projects have recently been carried out to study the ethology and mechanisms of psychiatric disorders. Although psychiatric disorders are part of the brain sciences, psychiatrists still diagnose them based on subjective experience rather than by gaining insights into the pathophysiology of the diseases. Deep learning (DL) methods have been increasingly applied to neuroimaging data to identify patients with psychiatric and neurological disorders. At present, psychiatric disorders are diagnosed based on symptoms and course of illness, according to the classifications in the Diagnostic and Statistical Manual of Mental Disorders.Given its ability to detect abstract and complex patterns, DL has been applied in neuroimaging studies of psychiatric and neurological disorders, which are characterised by subtle and diffuse alterations.Methods:Google Scholar, PubMed, and Web of Science databases were comprehensively searched from ۲۰۱۷ to ۲۰۲۳. The keywords used included psychiatric disorders, Deep learning, and neuroimaging. After removing irrelevant studies, finally ۱۰ studies that were compatible with the inclusion criteria were selected.Results:Although DL techniques have been explored extensively in various aspects of medical imaging, they are still in a relatively early stage, and most applications are still simple two- or three-classification problems. When studying fMRI from schizophrenia, psychotic bipolar disorder, schizoaffective disorder, and healthy individuals, the accuracy of a ۴-class classification reached ۴۶%, significantly above chance. The proposed deep classification and clustering framework is not only able to identify psychiatric disorders with high accuracy, but also interpret the correlation between brain networks and specific psychiatric disorders and reveal the relationship between them. Results suggest that deep learning of neuro imaging data is a promising tool for the classification of individual psychiatric patients.Conclusion(s):Deep learning provides a promising way to investigate a spectrum of similar disorders using neuroimaging-based measures. The combined development of psychiatric imaging and machine learning will be the trend and will become an indispensable tool for clinical diagnosis and treatment of psychiatric diseases in the future.