0069 DEVELOPMENT AND VALIDATION OF AN ALGORITHM FOR THE STUDY OF SLEEP USING A BIOMETRIC SHIRT IN YOUNG HEALTHY ADULTS

Abstract Introduction: Portable polysomnography systems are often too complex and encumbering for home sleep recordings. We assessed the feasibility of measuring sleep with a biometric shirt. Methods: Twenty healthy young adults (12 women, 8 men; 21.9 ± 2.0 years) were recorded in a sleep laboratory for two consecutive nights using standard polysomnography and a biometric shirt, simultaneously. Polysomnographic recordings were scored using standard methods. The biometric shirt had embedded electrocardiogram sensors, two respiratory inductance plethysmography bands, a 3-axis accelerometer and a detachable microcontroller performing signal acquisition, data processing and communication protocols. The shirt size was selected for each subject so that the signal was optimal. An algorithm was developed to classify the biometric shirt recordings into three vigilance states: wake, nonREM sleep and REM sleep. The algorithm was based on breathing rate and heart rate variability, body movement and included a correction for sleep onset and offset. The results from the two types of recordings were compared with percentages of agreement and kappa coefficients. Results: Five nights from four subjects were rejected due to recurrent signal artefacts caused by an ill-fitting or misplaced shirt. The overall mean percentage of agreement for 35 recording pairs was 77.55%. When NREM and REM sleep epochs were grouped together, the agreement was 90.7%. The overall kappa was 0.53. Removing breathing rate from the algorithm decreased kappa to 0.34 ± 0.13, whereas removing heart rate did not significantly modify it (0.54 ± 0.13). Five of the seven sleep variables were significantly correlated (sleep latency, total sleep time, %NREM and %REM sleep, the sleep period, wake time after sleep onset and sleep efficiency) while the minutes spent in NREM and of REM sleep did not. Conclusion: The findings of this pilot study indicate that a simple portable system using a biometric shirt can estimate reasonably well the general sleep pattern of young healthy adults. Support (If Any): Fondation Les Petits Trésors de l’Hôpital Rivière-des-Prairies, Montréal, QC Canada.