E-MDD: an evidence-guided explainable framework for EEG-based depression recognition
- 1College of Mechanical Engineering, Beijing Institute of Petrochemical Technology , ,
- 2College of Mechanical Engineering, Tiangong University , ,
Abstract
Introduction
An evidence-guided explainable framework, E-MDD, is developed to integrate high-dimensional EEG representations, neurophysiological biomarkers, and large language models for the recognition of EEG-based depression with concise and evidence-based textual explanations.Methods
The proposed E-MDD framework consists of three stages: classification, evidence construction, and evidence-guided inference. In the classification stage, two parallel feature extraction branches are employed to extract Electroencephalography latent features and neurophysiological biomarkers, which are subsequently fused and fed into a lightweight multilayer perceptron for the recognition of depression. In the evidence construction stage, the latent feature dimensions of the EEG are aligned with neurophysiological biomarkers, and the model's decision basis is identified to construct structured evidence supporting the classification results. In the evidence-guided inference stage, structured evidence and the proposed evidence-guided prompt are used to guide the large language model in generating concise textual explanations.Results
The experimental results on the Mumtaz dataset show that E-MDD achieves an AUC of 0.8472 and an F1-score of 0.8421. Meanwhile, the generated short-text explanations can effectively describe the basis of the model's decision-making for its predictions, demonstrating a high degree of faithfulness. The proposed framework also demonstrates consistent evidence-guided explanation performance across EEGNet and EEGConformer.Discussion
Overall, this study proposes a neurophysiologically interpretable framework for EEG-based deep learning models, which effectively enhances the interpretability of deep learning models for depression detection and provides a new perspective for understanding their decision-making mechanisms.
Related
- Physics-guided explainable machine learning for predicting temporal scour evolution around hybrid spur dikes under unsteady flood hydrographs
- Disentangling Context-Specific Alzheimer's Disease Progression Signatures through an Explainable AI Framework
- Grounding-Specific Effects of Valence in Word Recognition: Evidence from Pain-Related Language.
- A hybrid SE-ResNet50 deep learning framework for high-accuracy and explainable cotton leaf disease classification
- Nonlinear associations between comprehensive extreme climate hazard and regional grain supply performance in China: evidence from explainable XGBoost–SHAP models
BibTeX
@article{huang2026emdd,
title = {{E-MDD: an evidence-guided explainable framework for EEG-based depression recognition}},
author = {Junfen Huang and Zhuhuayang Zhang and Yingyu Cao and Zhenxi Zhao and Liangfei Niu},
journal = {Frontiers in digital health},
year = {2026},
month = oct,
doi = {10.3389/fdgth.2026.1950514},
url = {https://europepmc.org/article/PMC/PMC13717697},
note = {CC BY},
}
RIS
TY - JOUR
TI - E-MDD: an evidence-guided explainable framework for EEG-based depression recognition
AU - Junfen Huang
AU - Zhuhuayang Zhang
AU - Yingyu Cao
AU - Zhenxi Zhao
AU - Liangfei Niu
PY - 2026
DA - 2026/10/15
JO - Frontiers in digital health
DO - 10.3389/fdgth.2026.1950514
UR - https://europepmc.org/article/PMC/PMC13717697
AB - Introduction An evidence-guided explainable framework, E-MDD, is developed to integrate high-dimensional EEG representations, neurophysiological biomarkers, and large language models for the recognition of EEG-based depression with concise and evidence-based textual explanations.Methods The proposed E-MDD framework consists of three stages: classification, evidence construction, and evidence-guided inference. In the classification stage, two parallel feature extraction branches are employed to extract Electroencephalography latent features and neurophysiological biomarkers, which are subsequently fused and fed into a lightweight multilayer perceptron for the recognition of depression. In the evidence construction stage, the latent feature dimensions of the EEG are aligned with neurophysiological biomarkers, and the model's decision basis is identified to construct structured evidence supporting the classification results. In the evidence-guided inference stage, structured evidence and the proposed evidence-guided prompt are used to guide the large language model in generating concise textual explanations.Results The experimental results on the Mumtaz dataset show that E-MDD achieves an AUC of 0.8472 and an F1-score of 0.8421. Meanwhile, the generated short-text explanations can effectively describe the basis of the model's decision-making for its predictions, demonstrating a high degree of faithfulness. The proposed framework also demonstrates consistent evidence-guided explanation performance across EEGNet and EEGConformer.Discussion Overall, this study proposes a neurophysiologically interpretable framework for EEG-based deep learning models, which effectively enhances the interpretability of deep learning models for depression detection and provides a new perspective for understanding their decision-making mechanisms.
SN - 2673-253X
ER -