Research Progress on Artificial Intelligence and Brain–Computer Interfaces in the Emotion Recognition of Patients with Depression
Main Article Content
Keywords
artificial intelligence, brain-computer interface, depression, emotion recognition, deep learning
Abstract
Depression represents a significant public health concern on a global scale. Conventional diagnostic approaches rely on subjective scales, which have limitations, including low consistency and difficulty in early identification. In recent years, the integration of artificial intelligence (AI) and brain‒computer interface (BCI) technologies has led to the development of new solutions for mood recognition in patients with depression. This review examines emotion recognition in depression via multimodal data, including EEG signals, eye tracking, and facial expression analysis. This highlights the application and performance of deep learning methods, such as the EEGNet and LSTM-CNN parallel models. The study also discusses multimodal fusion techniques, including graph neural networks and dynamic weighted fusion. Research indicates that AI and BCI technologies enable objective quantification, real-time monitoring, and personalised intervention. These capabilities significantly increase the accuracy and robustness of depression recognition. Nevertheless, this field continues to face challenges, including ethical controversies and the limited generalizability of models. Future work should prioritise algorithmic improvements, facilitate clinical translation, and strengthen ethical frameworks to support broader implementation.
References
- Chen, X., Wang, Y. and Valdes-Sosa, P. A., (2023). Dynamic-weighted fusion of EEG and eye-tracking for depression detection: A bimodal deep learning approach. IEEE Transactions on Affective Computing, vol. 14, no. 2, pp. 1021–1034.
- Liu, Z., Zhang, H. and Zhang, R., (2024). Published. Contrastive learning for cross-modal depression recognition: An EEG-eye movement study. Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024), Seoul. IEEE, pp. 1-5.
- Song, X. W., (2023). Research on depression detection method based on EEG signal. Master's Thesis, Shandong University.
- Tan, F. D., Ma, X. L., Wang, Y. Y., Tao, T. T., Wang, G. Y. and Huang, S. H., (2024). Application of multi-modality based on artificial intelligence feature fusion in depression recognition. Modern Instruments & Medical Treatment, vol. 30, no. 1, pp. 58-63.
- Thibodeau, R., Jorgensen, R. S. and Kim, S., (2023). Frontal EEG alpha asymmetry as a diagnostic marker of depression: A meta-analytic update. Neuroscience & Biobehavioral Reviews, vol. 145, p. 105002.
- Xie, W., Wang, C., Lin, Z., Luo, X., Chen, W., Xu, M., Liang, L., Liu, X., Wang, Y. and Luo, H., (2022). Multimodal fusion diagnosis of depression and anxiety based on CNN-LSTM model. Computerized Medical Imaging and Graphics, vol. 102, p. 102128.
- Yang, Q. R., Fang, W. S., Xu, X. Y. and Xu, Z. Y., (2025). Emotion recognition based on LSTM-CNN parallel neural networks. Journal of Luoyang Institute of Science and Technology(Natural Science Edition), vol. 35, no. 2, pp. 69-73.
- Zhang, Y. C., Li, X. F. and Wang, Z. Q., (2022). EEG-eye movement depression emotion recognition based on graph neural network. Journal of Computer Research and Development, vol. 59, no. 8, pp. 1801–1812.
- Zhang, Y. H., Nie, Y. M., Sun, B., He, J. and Yang, B., (2018). A method of depression evaluation based on decision level fusion of deep forest with multimodality data. Journal of Beijing Normal University(Natural Science), vol. 54, no. 5, pp. 606-611.
- Zhao, J., Zhou, L. Y. and Wu, M. Q., (2023). Artificial intelligence-based auxiliary diagnosis method for depression. Journal of Northwest University (Natural Science Edition), vol. 53, no. 3, pp. 325–335.
Article Sidebar
This work is licensed under a Creative Commons Attribution 4.0 International License.