Presentation Information
[SY-66-02]Diagnosing and Treating Major Depressive Disorder Using EEG-Based Machine Learning
*Hsin-An Chang1, Yi-Hung Liu2 (1.Department of Psychiatry, Tri-Service General Hospital, National Defense Medical Center, Taipei(Taiwan), 2.Institute of Electrical and Control Engineering, National Yang Ming Chiao Tung University, Hsinchu(Taiwan))
Keywords:
Major depressive disorder,EEG,Machine learning
Electroencephalography is a widely used research and clinical tool to monitor and record the electrical activity of the brain – the electroencephalogram (EEG). Machine learning algorithms have been developed to extract features from the EEG to identify various brain states from different neuropsychiatric disorders. Major depressive disorder (MDD) is a leading mental disorder worldwide. According to the World Health Organization, the annual global economic impact of depression is estimated at $1 trillion and is projected to be the leading cause of disability by 2020. Nowadays, the role of artificial intelligence in efforts to diagnose and treat MDD is getting more and more important. A growing body of research aims to better predict, diagnose, and treat MDD by using EEG-based machine learning as a potential solution. Our research team aims to explore the role of EEG-based machine learning in supporting depression diagnosis and treatment response prediction. We previously used EEG-based machine learning model to classify MDD patients versus healthy controls with acceptable accuracy. We subsequently used the combination of EEG-based machine learning plus self-reported depression severity to predict MDD patients with suicidal risks. In real-world observational studies, we tested the performance of the models of machine learning trained from the resting-state EEG data at baseline to predict treatment response to either 8–week antidepressant treatment in MDD patients or 30-session repetitive transcranial magnetic stimulation (rTMS) in treatment-resistant MDD patients. The results showed that specific machine learning classifiers can effectively predict treatment response in these patients. EEG-based machine learning shows substantial promise in the diagnosis and management of depression. However, the applications of EEG-based machine learning require further validation before they can be relied upon as diagnostic tools or a biomarker to predict treatment response.