R H Y Lee, S Vukovic, V E M Lucci, J T Lando, V E Claydon
Ambulatory autonomic and sleep monitoring in real-world settings remains challenging. The Astroskin biometric system may offer a practical solution. Here, we evaluate the utility of Astroskin as a wearable research device compared to common industry standard approaches. We assessed the Astroskin under static and dynamic conditions, measuring peripheral blood oxygen saturation (SpO2) during hypobaric hypoxia; skin temperature and estimated core temperature under insulated thermal stress; heart rate (HR), systolic blood pressure (SBP), and respiratory rate and depth during Valsalva maneuvers and positional changes, and sleep staging during at-home sleep. Data were compared using Pearson's correlations, Bland-Altman plots for accuracy and precision, and two one-sided t-tests for equivalency. SpO2 (r = 0.93, p < 0.01; bias -2.95 ± 2.7%), skin temperature (r = 1.0, p < 0.01; bias -0.30 ± 0.43 °C), and HR (r = 0.71, p < 0.01; bias -1.77 ± 7.22 bpm) are strongly correlated, highly accurate, and equivalent to within 5% of the industry comparators. Estimated core temperatures also achieved high agreement with rectal temperature (r = 0.81, p < 0.01; bias +0.34 ± 0.2 °C; 1% equivalency). Respiratory parameters are optimized when derived directly from band stretch. SBP measurements approach industry standards with high accuracy and equivalency, but low precision (r = 0.56, p < 0.01; bias -2.13 ± 11.57 mmHg; 3% equivalency). Despite appropriate sleep detection (100% success), Astroskin's sleep staging algorithm was unpredictable (41% success). When staging was available, Astroskin differentiated sleep from wake with equivalency within 10%, including sensitive detection of arousals. The Astroskin is suitable for ambulatory monitoring, with accuracy optimized when systematic biases are corrected. Use for sleep assessment requires improved sleep staging success. Astroskin's practical, continuous, wearable approach underscores its potential for diverse applications.