An Automated System to Assist Clinicians in the Detection of Insomnia using Wearable Sensors

Abstract

The diagnosis and treatment of insomnia is time-consuming and complicated due to large variability in patient symptoms. We thus developed an ambulatory monitoring system consisting of 1) a mobile monitoring application comprising of a sleep diary and wearable physiological sensors; 2) a technological aide that records and monitors patient data for a differential diagnosis of insomnia; and 3) a central server, where novel signal processing algorithms are used to detect relevant insomnia PSG features and clinician reports generated. When tested with clinical data, our algorithms detected insomnia specific features and events with accuracies of 81-93%, indicating that remote monitoring system of insomnia is feasible.