{"title":"E-DANN: An Enhanced Domain Adaptation Network for Audio-EEG Feature Decoupling in Explainable Depression Recognition","authors":"Qinglin Zhao;Hua Jiang;Zhongqing Wu;Lixin Zhang;Kunbo Cui;Kai Zheng;Jingyu Liu;Ran Cai;Mingqi Zhao;Fuze Tian;Bin Hu","doi":"10.1109/TNSRE.2025.3608181","DOIUrl":null,"url":null,"abstract":"Given the significant global health burden caused by depression, numerous studies have utilized artificial intelligence techniques to objectively and automatically detect depression. However, existing research primarily focuses on improving the accuracy of depression recognition while overlooking the explainability of detection models and the evaluation of feature importance. In this paper, we propose a novel framework named Enhanced Domain Adversarial Neural Network (E-DANN) for depression detection. First, we extract joint features that combine audio-specific physical properties and electroencephalogram (EEG) responses to construct a multimodal feature space, which facilitates the understanding of how EEG signals dynamically respond to the physical properties of audio. Next, we employ the feature decoupling framework of E-DANN, which separates the extracted feature space into shared features and private features through adversarial training. The decoupled private features are then utilized for the binary classification of depression. Our experimental results validate the effectiveness of the proposed framework, which achieves accurate classification of normal controls and individuals with depression (accuracy: <inline-formula> <tex-math>$92.83~\\pm ~4.38$ </tex-math></inline-formula>%, specificity: <inline-formula> <tex-math>$93.56~\\pm ~7.25$ </tex-math></inline-formula>%, sensitivity: <inline-formula> <tex-math>$91.61~\\pm ~6.87$ </tex-math></inline-formula>%, and F1 score: <inline-formula> <tex-math>$91.81~\\pm ~4.52$ </tex-math></inline-formula>%). Furthermore, we employ an Explainable Artificial Intelligence (XAI) approach to hierarchically visualize feature importance and elucidate complex feature interaction patterns. In summary, this study provides a theoretical foundation for developing explainable diagnostic tools for depression and contributes to improving the clinical trustworthiness of AI-assisted diagnostic systems.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"33 ","pages":"3647-3661"},"PeriodicalIF":5.2000,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11156140","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/11156140/","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Given the significant global health burden caused by depression, numerous studies have utilized artificial intelligence techniques to objectively and automatically detect depression. However, existing research primarily focuses on improving the accuracy of depression recognition while overlooking the explainability of detection models and the evaluation of feature importance. In this paper, we propose a novel framework named Enhanced Domain Adversarial Neural Network (E-DANN) for depression detection. First, we extract joint features that combine audio-specific physical properties and electroencephalogram (EEG) responses to construct a multimodal feature space, which facilitates the understanding of how EEG signals dynamically respond to the physical properties of audio. Next, we employ the feature decoupling framework of E-DANN, which separates the extracted feature space into shared features and private features through adversarial training. The decoupled private features are then utilized for the binary classification of depression. Our experimental results validate the effectiveness of the proposed framework, which achieves accurate classification of normal controls and individuals with depression (accuracy: $92.83~\pm ~4.38$ %, specificity: $93.56~\pm ~7.25$ %, sensitivity: $91.61~\pm ~6.87$ %, and F1 score: $91.81~\pm ~4.52$ %). Furthermore, we employ an Explainable Artificial Intelligence (XAI) approach to hierarchically visualize feature importance and elucidate complex feature interaction patterns. In summary, this study provides a theoretical foundation for developing explainable diagnostic tools for depression and contributes to improving the clinical trustworthiness of AI-assisted diagnostic systems.
期刊介绍:
Rehabilitative and neural aspects of biomedical engineering, including functional electrical stimulation, acoustic dynamics, human performance measurement and analysis, nerve stimulation, electromyography, motor control and stimulation; and hardware and software applications for rehabilitation engineering and assistive devices.