Beyond Augmentation: Score-Guided Pathological Prior for EEG-based Depression Detection

2026-05-29

AI Summary

This study addresses the performance limitations of EEG-based depression detection models caused by small sample sizes. The authors propose a Score-Guided Classification framework that eschews conventional data augmentation and synthetic sample generation. Instead, it employs an unsupervised generative network to model the degree of pathological abnormality in each sample, which serves as a “pathological prior” fused with deep features to guide decision boundary formation. Additionally, a cross-channel spatial adaptation module is introduced to mitigate channel heterogeneity across multi-center datasets. Experiments on the Mumtaz2016 and MODMA datasets demonstrate that the proposed method significantly enhances both detection accuracy and generalization capability under zero data augmentation, establishing a novel and effective learning paradigm that operates without synthetic data.

Abstract

Deep learning-based Major Depressive Disorder (MDD) detection using Electroencephalography (EEG) is fundamentally constrained by the "small-sample dilemma." Prevailing generative data augmentation methods not only incur heavy computational overhead but also risk introducing synthetic noise, thereby blurring classification boundaries. To challenge the traditional "data quantity first" convention, we propose a novel framework "Beyond Augmentation": Score-Guided Classification (SGC). SGC does not synthesize pseudo-samples; instead, it utilizes an unsupervised generative network architecture to model the structural and statistical anomaly degrees of samples, serving as the core "Pathological Prior". This prior, after robust normalization, is explicitly fused with deep feature representations, thereby precisely guiding the classifier's decision boundary. Furthermore, to dynamically adapt to varying channel configurations, we propose a Cross-Channel Spatial Adaptation module, utilizing a spatial mapping mechanism to effectively resolve the hardware heterogeneity of mismatched channels in multi-center datasets. Extensive experiments on the Mumtaz2016 and high-density MODMA datasets demonstrate the effectiveness and exceptional generalizability of our method under the challenging "zero data augmentation" setting and at "zero sample synthesis cost". Keywords: Electroencephalography (EEG), Depression Detection, Anomaly Score, Diffusion Models, Few-Shot Learning

Problem

• Electroencephalography (EEG)
• Depression Detection
• Small-sample dilemma
• Data Augmentation
• Few-Shot Learning

Innovation

• Pathological Prior
• Score-Guided Classification
• Anomaly Score
• Zero Data Augmentation
• Cross-Channel Spatial Adaptation