Predicting depressed patients with suicidal ideation from ECG recordings

• Published in: Medical & biological engineering & computing Vol. 55; no. 5; pp. 793 - 805
• Main Authors: Khandoker, A. H., Luthra, V., Abouallaban, Y., Saha, S., Ahmed, K. I., Mostafa, R., Chowdhury, N., Jelinek, H. F.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.05.2017; Springer Nature B.V
• Subjects: Adult; Age; Aged; Aged, 80 and over; Autonomic nervous system; Biomedical and Life Sciences; Biomedical Engineering and Bioengineering; Biomedicine; Case-Control Studies; Classification; Computer Applications; Demographics; Depressive Disorder, Major - physiopathology; Depressive Disorder, Major - psychology; EKG; Electrocardiography; Electrocardiography - methods; Entropy; Female; Heart rate; Human Physiology; Humans; Imaging; Male; Mental depression; Modulation; Original Article; Pilot Projects; Preprocessing; Radiology; Regression analysis; Risk factors; Self destructive behavior; Suicidal behavior; Suicidal Ideation; Classification and regression tree; Suicidal ideation; Heart rate variability; Major depressive disorder; Tone–entropy analysis
• ISSN: 0140-0118; 1741-0444
• DOI: 10.1007/s11517-016-1557-y

Globally suicidal behavior is the third most common cause of death among patients with major depressive disorder (MDD). This study presents multi-lag tone–entropy (T–E) analysis of heart rate variability (HRV) as a screening tool for identifying MDD patients with suicidal ideation. Sixty-one ECG recordings (10 min) were acquired and analyzed from control subjects (29 CONT), 16 MDD subjects with (MDDSI+) and 16 without suicidal ideation (MDDSI−). After ECG preprocessing, tone and entropy values were calculated for multiple lags ( m: 1–10). The MDDSI+ group was found to have a higher mean tone value compared to that of the MDDSI− group for lags 1–8, whereas the mean entropy value was lower in MDDSI+ than that in CONT group at all lags (1–10). Leave-one-out cross-validation tests, using a classification and regression tree (CART), obtained 94.83 % accuracy in predicting MDDSI+ subjects by using a combination of tone and entropy values at all lags and including demographic factors (age, BMI and waist circumference) compared to results with time and frequency domain HRV analysis. The results of this pilot study demonstrate the usefulness of multi-lag T–E analysis in identifying MDD patients with suicidal ideation and highlight the change in autonomic nervous system modulation of the heart rate associated with depression and suicidal ideation.