Development of a Depression Screening System Using Heart Rate Variability and Cerebral Blood Flow ∼ An Algorithm Based on Multiple Mental Load Tasks ∼

Abstract

In order to achieve an objective major depressive disorder (MDD) screening, we developed an MDD screening system via linear discriminant analysis (LDA). Using transient responses induced by multiple mental tasks that random number generation (RNG) task and verbal fluency (VF) task, the system discriminates MDD patients from normal subjects using heart rate variability (HRV) indices and cerebral blood flow determined by near-infrared spectroscopy (NIRS). The screening was conducted on twelve MDD patients (32±15 years) and 30 control subjects (37±17 years). The mental work load procedures are composed of "Pre-task rest", "Undergoing task", and "After-work rest" sections. When we adopted only RNG task, the system indicated 83.3% sensitivity and 93.8% specificity and with only VF task, the sensitivity and specificity were 75% and 90%, respectively. However using multiple tasks, we were able to obtain 100% sensitivity and 100% specificity with our proposed system.