Force-Sensing-Based Unobtrusive System for Awakening and Respiration Rate Analysis During Sleep

In this paper, we designed an unobtrusive strip-type sleep monitoring system including a force sensor and algorithms for the detection of awakening and respiration rate. The developed system is installed on a mattress across the user's chest in order to measure the trunk movement induced by respiration or body motion. Thirty subjects' clinical data, which contain signals that were measured simultaneously using the developed system and a polysomnograph (PSG), are used to validate the accuracy of the developed system. The awakening detection, accuracy, sensitivity, specificity, and Kappa coefficient of the designed system were 79.4%, 48.3%, 86.9%, and 0.6, respectively, and the average sleep efficiency was 82.3% in the developed system and 84.0% in the PSG with a root-mean-square error of 17.7%. In terms of respiration rate detection, it was shown that the respiration rate measured with the developed system can trace the PSG-measured respiration rate. The accuracy to detect respiration rates within 1-3 breaths per minute error were 76.4%, 97.2%, and 99.4%, respectively.