Practical Strategies for Longitudinal in-Home Sleep Tracking in Diverse Populations

Abstract

Longitudinal monitoring of sleep metrics, night-to-night sleep quality, and weekly sleep patterns is important for enabling healthy aging and, more importantly, understanding and mitigating cognitive decline with aging. Recent work has shown that longitudinal changes in selected sleep metrics, such as total sleep time, time in the rapid eye movement (REM) versus non-REM sleep phases, and non-REM slow wave activity, is closely correlated with decline in cognitive function in older populations. Further, evidence also suggests that sleep disturbance in middle-aged and older populations can contribute to cognitive decline and heighten the risk of Alzheimer's Disease (AD) by increasing beta-amyloid burden. In older, at-risk populations, persistent sleep disturbances and/or shifts in sleep patterns also precede an AD diagnosis and may even appear years before cognitive decline, indicating that clinically-accurate, longitudinal sleep tracking can allow for on-time therapeutics and interventions before onset of AD symptoms.However, clinical-grade wearable sleep monitoring is a challenging problem since it requires simultaneous monitoring of brain activity, eye movement, muscle activity, cardio-respiratory features, and gross body movements. This requires multiple sensors to be worn at different locations as well as uncomfortable adhesives and discrete electronic components to be placed on the head. As a result, existing wearables either compromise comfort or compromise accuracy in tracking sleep variables.I will describe our lab's recent work in developing a suite of fabric-based sensors and garment-integrated sensing systems for in-home sleep tracking, focusing particularly on our efforts at enabling clinically-accurate sleep phase recognition. Our best-performing solution is PhyMask, an all-textile sleep monitoring solution that is practical and comfortable for continuous use and that acquires all signals of interest to sleep solely using comfortable textile sensors placed on the head. We show that PhyMask can be used to accurately measure all the signals required for precise sleep stage tracking and to extract advanced sleep markers such as spindles and K-complexes robustly in the real-world setting. We validate PhyMask against polysomnography and show that it significantly outperforms other commercially-available sleep tracking wearables, including the Fitbit wristband, the Oura Ring, and the Muse, Phillips SmartSleep and Dreem headbands. Figure 1