A Generalized Depression Recognition Framework Based on Cross-Center and Cross-Task Eeg Signals

Abstract

This study developed a generalized framework for automatic depression recognition using a Dempster Shafer Theory-based Classification Fusion (DSTCF) method. We collected resting-state electroencephalography (EEG) data using 128 electrodes from 24 Major Depressive Disorder (MDD) patients and 29 Normal Controls (NC), as well as additional data from 47 MDD and 47 NC in both resting-state and dot-probe task conditions. By extracting eight linear and three nonlinear features and applying three feature selection methods, we conducted cross-center recognition and validation. Cross-center validation using the DSTCF method showed that after adopting the Relief method, EEG data in the beta frequency band achieved optimal performance. The best average accuracy was 96.18% across three training datasets and 67.82% across six validation datasets, representing improvements of 7% and 8%, respectively, compared to traditional methods. Furthermore, Ppmean and activity features demonstrated high discriminative capability and significant correlation with depressive levels. Based on these two features, abnormal parieto-occipital lobe activation served as a task-independent feature for MDD identification. Additionally, abnormal activation in the frontal lobe’s beta frequency band and the temporal lobe’s theta frequency band can effectively distinguish MDD from NC under resting and task conditions, respectively. These findings not only help in understanding the atypical neural mechanisms of depression but also provide reliable EEG biomarkers for its identification and diagnosis.

Keywords: EEGMajor depressive disorderCross-centerCross-taskDempster-Shafer TheoryFeature selectionDecision-level fusion