Performance Evaluation of the Circadia Contactless Breathing Monitor and Sleep Analysis Algorithm for Sleep Stage Classification

Although polysomnography (PSG) remains the gold standard for studying sleep in the lab, the development of wearable and `nearable' non-EEG based sleep monitors has the potential to make long-term sleep monitoring in a home environment possible. However, validation of these novel technologies against PSG is required. The current study aims to evaluate the sleep staging performance of the radar-based Circadia Contactless Breathing Monitor (model C100) and proprietary Sleep Analysis Algorithm, both in a home and sleep lab environment, on cohorts of healthy sleepers. The C100 device was initially used to record 17 nights of sleep data from 9 participants alongside PSG, with a subsequent 24 nights of PSG data for validation purposes. Respiration and body movement features were extracted from sensor data, and a machine learning algorithm was developed to perform sleep stage prediction. The algorithm was trained using PSG data obtained in the initial dataset (n=17), and validated using leave- one-subject-out cross-validation. An epoch-by-epoch recall (true positive rate) of 75.0 %, 59.9 %, 74.8 % and 57.1 %, was found for `Deep', `Light', `REM' and `Wake' respectively. Highly similar results were obtained in the independent validation dataset (n=24), indicating robustness of results and generalizability of the sleep staging model, at least in the healthy population. The device was found to outperform both a consumer and medical grade wrist-worn monitoring device (Fitbit Alta HR and Philips Respironics Actiwatch) on sleep metric estimation accuracy. These results indicate that the developed non-contact monitor forms a viable alternative to existing clinically used wrist-worn methods, and that longitudinal monitoring of sleep stages in a home environment becomes feasible.