Modernizing MDD Diagnosis using Deep Learning from EEG Data

Abstract

Major depressive disorder (MDD) is a widespread illness significantly impacting individuals’ quality of life. Its diagnosis through Electroencephalogram (EEG) has long been studied in mental health research. Recent advancements in deep learning present a promising pathway for enhancing MDD diagnosis through EEGs. This study integrates state-of-the-art deep learning techniques, including ConvNext and Transformers architectures, into MDD prediction models. Results demonstrate ConvNext models’ robustness and efficiency, in terms of precision and specificity, while Transformer models exhibit high recall and sensitivity for diagnosing MDD from incomplete studies.