Asleep Adults' Breathing Patterns via Data Mining of Electromyograms

Abstract

A study used Machine Learning to identify potential sleep disturbances by analyzing abdominal electromyograms and breathing patterns in sleeping adults. The median and peak values of envelopes of electromyograms have been used as diagnostic signs of breathing patterns. The report suggests using visual classifiers, like decision trees, to make the diagnostic process more accessible for clinicians unfamiliar with Machine Learning. The duly preprocessed abdominal electromyograms allow the forecast of the breathing pattern by decision trees with a precision of up to 0.848.