Predicting depressed patients with suicidal ideation from ECG recordings

Area of Science:

• Cardiology
• Psychiatry
• Biomedical Engineering

Background:

• Suicidal behavior is a leading cause of death in major depressive disorder (MDD) patients.
• Accurate identification of suicidal ideation in MDD is crucial for timely intervention.
• Current screening methods may benefit from advanced physiological markers.

Purpose of the Study:

• To evaluate multi-lag tone-entropy (T-E) analysis of heart rate variability (HRV) as a screening tool for suicidal ideation in MDD patients.
• To compare the efficacy of T-E analysis with traditional HRV metrics and demographic factors.

Main Methods:

• Acquired 10-minute ECG recordings from 61 participants: 29 controls, 16 MDD with suicidal ideation (MDDSI+), and 16 MDD without suicidal ideation (MDDSI-).
• Calculated multi-lag tone and entropy values from preprocessed ECG data (lags m: 1-10).
• Employed classification and regression tree (CART) with leave-one-out cross-validation.

Main Results:

• MDDSI+ group exhibited higher mean tone values (lags 1-8) and lower mean entropy values (lags 1-10) compared to MDDSI- and control groups, respectively.
• A combination of T-E values and demographic factors achieved 94.83% accuracy in predicting MDDSI+.
• T-E analysis outperformed traditional time and frequency domain HRV analyses.

Conclusions:

• Multi-lag T-E analysis of HRV is a promising, accurate tool for identifying MDD patients with suicidal ideation.
• This method highlights alterations in autonomic nervous system modulation linked to depression and suicidal ideation.
• Further validation is warranted for clinical application in suicide risk assessment.