Cognitive Depression Detection Cyber-Medical System Based on EEG Analysis and Deep Learning Approaches.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Biomedical Engineering

Background:

• Depression significantly impacts quality of life and productivity.
• Current depression detection methods, primarily subjective scales, lack speed and objectivity.
• Identifying objective biomarkers for depression is crucial for timely intervention.

Purpose of the Study:

• To empirically identify brainwave stimulation feedback electrode points and brain regions associated with depression.
• To develop and evaluate deep neural network models for objective depression assessment.
• To explore the potential for an auxiliary system for rapid depression detection and emotional self-management.

Main Methods:

• Collected brainwave data using mood-induction procedures.
• Applied signal processing techniques (Fourier and wavelet transforms) to analyze brainwave bands (α and θ).
• Developed and compared 8 depression assessment models using various deep neural network architectures (MLP, DNN, DBN, LSTM).

Main Results:

• The front (Fp1, Fp2) and occipital lobes (O1, O2) were identified as key brain regions involved in depressive emotions.
• Deep neural network models demonstrated superior and stable performance in depression assessment.
• Specific brainwave bands (α, θ) and signal characteristics were significantly affected in individuals experiencing depressive states.

Conclusions:

• Deep neural networks offer a robust foundation for an objective, rapid depression assessment system.
• This technology can facilitate early detection, autonomous emotional management, and personalized treatment strategies.
• Identifying individual abnormalities during low mood stages can guide targeted relief methods, potentially reducing depression incidence.