Auxiliary identification of depression patients using interpretable machine learning models based on heart rate...

Area of Science:

• Cardiology
• Psychiatry
• Machine Learning

Background:

• Depression is a significant global health concern with high incidence and disability.
• Accurate and efficient diagnosis is crucial for timely intervention.

Purpose of the Study:

• To explore the association between heart rate variability (HRV) and depression.
• To establish and validate machine learning models for auxiliary depression diagnosis using HRV.

Main Methods:

• Utilized data from 465 outpatients, randomly split into training (70%) and test (30%) sets.
• Developed and compared four machine learning models: Logistic Regression, Support Vector Machine, Random Forest, and XGBoost.
• Evaluated model performance using ROC curves, calibration curves, and decision curve analysis, with SHAP for feature importance.

Main Results:

• The XGBoost model demonstrated superior performance with Area Under the Curve (AUC) values of 0.92 in the training set and 0.82 in the test set.
• Top predictors identified by SHAP included age, heart rate, Standard Deviation of NN intervals (SDNN), two nonlinear HRV parameters, and sex.
• The XGBoost model exhibited excellent predictive efficacy and clinical utility.

Conclusions:

• HRV is a feasible potential biomarker for depression.
• The HRV-based machine learning model serves as a quantitative auxiliary diagnostic tool for clinicians.
• This tool can enhance diagnostic accuracy and efficiency, aiding in depression monitoring and prevention.