A systematic literature review and analysis of deep learning algorithms in mental disorders

Abstract

Introduction: Mental disorders are the main cause of mortality and morbidity worldwide. Deep learning offers a considerable promise for mental health diagnosis and risk assessment. The current study considered the potential application of deep learning methods in mental disorders. Method: Four databases were reviewed between 2000 and February 2023, based on the PRISMA methodology. A total of 1339 papers was recognized and screened for their relevance to the use of deep learning algorithms in mental disease; 85 pertinent studies were identified and categorized based on several dimensions, such as subspecialty, deep learning methods, data sources, study limitations, and future directions. Result: The obtained result revealed that deep learning in mental health is vastly used for depression and mood recognition analysis. The Convolutional Neural Network (CNN) is a prominent method applied in selected studies. Conclusion: The results of this study may motivate further research on the use of deep learning in mental disorders and future directions for this promising technology.

Keywords

• Deep learning
• Artificial neural networks
• Mental health
• Mental disorders
• Literature review