0090 Performance of a Multisensor Ring to Evaluate Sleep: In-Lab Evaluation Relative to PSG and Actigraphy: Importance of Generalized Versus Personalized Scoring

Abstract Introduction Multisensor sleep wearable devices have demonstrated utility for research and relative accuracy discerning sleep-wake patterns at home in the laboratory. Additional sensors more complex scoring algorithms may improve ability of wearables to assess health. Methods Thirty-six healthy adults completed assessment while wearing experimental device (Happy Ring), as well Philips Actiwatch, Fitbit, Oura, Whoop devices. Evaluations occurred laboratory (Alice 6 polysomnogram). The Happy Ring includes accelerometry, photoplethysmography, electrodermal activity, skin temperature. Epoch-by-epoch analyses compared lab polysomnography, other sleep-tracking Scoring was accomplished using two machine-learning-derived algorithms: a “generalized” algorithm which static applied all users (like those used devices) “personalized” where parameters are personalized, dynamic, change based on data collected across different parts night sleep. Results Compared in-lab generalized from good sensitivity (94%) specificity (70%). personalized also performed with (93%) (83%). Other sensitivity, ranging 89% (Fitbit) 94% (Actiwatch); however, variable, 19% (Actiwatch) 54% (Whoop). Overall 91% 92% 88% 86% Whoop, 84% 85% Actiwatch. Measurement stage 67%, 85%, light, deep, REM sleep, respectively, algorithm. For algorithm, stages were 81%, 95%, 92%, deep respectively. Post-hoc showed that better than modalities (p<0.001). Kappa scores 0.45 0.68 0.32 Oura Ring, Strap, 0.37 Fitbit wristband. Conclusion multisensory ring detection approach outperformed others, representing potential innovation improving accuracy. Support (If Any) Dr. Grandner is supported by R01DA051321 R01MD011600. This work Health, Inc.