Event-Related NIRS and EEG Analysis for Mental Stress Monitoring

Abstract

Mental disorders caused by chronic stress are difficult to identify, and colleagues in the work environment may suddenly report symptoms. Social barriers exist including the financial cost of medical services and the lack of a perceived need for treatment even if potential patients have a desire to receive mental healthcare. Self-report inventories such as the Beck Depression Inventory (BDI-II) and State-Trait Anxiety Inventory (STAI) can assess the emotional valence for mental health assessment, but medical expertise may be required for interpretation of the results. Contingency plans for clinical supervision and referral sources are necessary for sufficient mental healthcare using self-report inventories. On the other hand, the laterality index at rest (LIR) has been proposed for evaluation of the mental stress level from near-infrared spectroscopy (NIRS) data in the prefrontal cortex in the resting state. However, the potential for long-term monitoring has not been investigated with sufficient evaluation results. In this study, feature values were extracted from both NIRS and EEG signals each week for 10 weeks in four young participants with an average BDI-II score of 17.7, i.e., indicative of mild depression. Temporal changes in LIR and heart rate (HR) were compared with STAI-Y1 and BDI-II scores. We found cross-correlations between the time series of LIR and STAI-Y1 within one-week delay. In addition, the time series of LIR was also correlated with BDI-II with one-week delay. Importantly, by annotating the larger changes in LIR and HR on daily life events, the changes in LIR and HR were different depending on the type of life event that affected these moods.