Related Experiment Video
Updated: Aug 6, 2025
05:19
Author Spotlight: Therapeutic Benefit of Closed-Loop Deep Brain Stimulation in Depression Treatment
Published on: July 7, 2023
2.4K
Cross-subject classification of depression by using multiparadigm EEG feature fusion
Jianli Yang1, Zhen Zhang2, Zhiyu Fu2
1College of Electronic Information and Engineering, Hebei University, Baoding 071002, China; Key Laboratory of Digital Medical Engineering of Heibei Province, Baoding 071002, China.
Computer Methods and Programs in Biomedicine
|March 21, 2023
Summary
This study enhances depression classification using electroencephalogram (EEG) signals by fusing data from eyes-open and eyes-closed states. Multiparadigm feature concatenation with SVM achieved 94.03% accuracy, improving diagnosis.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Computational Psychiatry
Background:
- Electroencephalogram (EEG) signals present challenges in depression classification due to non-stationarity, complexity, and individual variability.
- Existing methods struggle to fully capture the nuances of EEG for accurate depression diagnosis.
Purpose of the Study:
- To develop an effective method for depression classification by addressing EEG signal complexities.
- To improve classification accuracy by integrating data from different resting-state paradigms (eyes open and eyes closed).
Main Methods:
- Extracted Lempel-Ziv complexity feature matrices from resting-state EEG under eyes-open and eyes-closed conditions.
- Employed topographical brain mapping and statistical analysis to assess paradigm significance.
- Implemented and compared linear combination and concatenation fusion methods for feature matrices.
- Utilized Support Vector Machine (SVM), K-nearest neighbor, and decision tree classifiers.
Main Results:
- The highest single-paradigm accuracy was 86.58% with eyes-open EEG.
- Multiparadigm feature concatenation outperformed linear combination.
- The SVM classifier with concatenated features achieved a peak accuracy of 94.03%.
Conclusions:
- Multiparadigm feature fusion significantly enhances depression classification accuracy.
- Eyes-open and eyes-closed EEG data provide complementary information beneficial for cross-subject depression classification.
- This approach offers novel strategies for clinical depression diagnosis using EEG.
