Multi-sources data analysis with sympatho-vagal balance estimation toward early bruxism episodes detection.

Abstract

Sleep bruxism events detection system is presented, based on integrated, synchronized on-line analysis of EMG signal, heart rave variability (HRV) obtained from ECG recordings as well as sympatho-vagal balance estimated in real time as an possible early indicator of upcoming bruxism episodes. As an relative reliable alternative for very complex systems, only for clinical environment usage with audio and video recordings a pilot study toward elaboration of compact, comfortable for home usage device with early bruxism detection algorithms was carried out, preliminary tested on 10h sleeping registrations from group of 12 patients, clinically characterized by experts as Bruxers. As a result a set of decision rules regarding simultaneous monotonic increase of heart rate with significant increase of EMG signal amplitude during bruxism episode was elaborated. But a most promising observation, which can be useful for earlier prediction of upcoming bruxism episode seems to be a monotonic increase of LF/HF ratio in HRV power spectrum components, expressing sympatho-vagal balance of autonomous nervous system, which according to our assumptions take basic low level role in bruxism phenomena trigger and control.