Complexity Index From a Personalized Wearable Monitoring System for Assessing Remission in Mental Health

This study discusses a personalized wearable monitoring system, which provides information and communication technologies to patients with mental disorders and physicians managing such diseases. The system, hereinafter called the PSYCHE system, is mainly comprised of a comfortable t-shirt with embedded sensors, such as textile electrodes, to monitor electrocardiogram-heart rate variability (HRV) series, piezoresistive sensors for respiration activity, and triaxial accelerometers for activity recognition. Moreover, on the patient-side, the PSYCHE system uses a smartphone-based interactive platform for electronic mood agenda and clinical scale administration, whereas on the physician-side provides data visualization and support to clinical decision. The smartphone collects the physiological and behavioral data and sends the information out to a centralized server for further processing. In this study, we present experimental results gathered from ten bipolar patients, wearing the PSYCHE system, with severe symptoms who exhibited mood states among depression (DP), hypomania(HM), mixed state (MX), and euthymia (EU), i.e., the good affective balance. In analyzing more than 400 h of cardiovascular dynamics, we found that patients experiencing mood transitions from a pathological mood state (HM, DP, or MX-where depressive and hypomanic symptoms are simultaneously present) to EU can be characterized through a commonly used measure of entropy. In particular, the SampEn estimated on long-term HRV series increases according to the patients' clinical improvement. These results are in agreement with the current literature reporting on the complexity dynamics of physiological systems and provides a promising and viable support to clinical decision in order to improve the diagnosis and management of psychiatric disorders.