WaaS architecture-driven depressive mood status quantitative analysis based on forehead EEG and self-rating tool

Area of Science:

• Neuroscience
• Computational Psychiatry
• Digital Health

Background:

• Objective depression evaluation is a growing research area, yet a pervasive, quantitative approach remains underdeveloped.
• Existing methods lack objective, real-world applicability for continuous depression monitoring.

Purpose of the Study:

• To propose a quantitative analysis method for assessing depressive mood status.
• To integrate forehead electroencephalograph (EEG) data with electronic diary logs for objective depression evaluation.

Main Methods:

• Developed a quantitative analysis method leveraging Wisdom as a Service architecture.
• Utilized forehead EEG features and quantitative log for mental state (Q-Log) data.
• Employed a random forest regression algorithm for model training.

Main Results:

• The quantitative model demonstrated a moderate positive correlation (average coefficient = 0.6556, P < 0.01) with Q-Log data using Leave-One-Participant-Out Cross-Validation.
• An exemplary knowledge-sharing application was developed using ontology and Jena inference for data annotation and result interpretation.

Conclusions:

• Combining physiological sensor data (EEG) with psychological self-rating data (Q-Log) offers novel insights.
• This integrated approach facilitates pervasive and objective depression evaluation in daily life.