A Multi-sensor Data Fusion Approach for Sleep Apnea Monitoring using Neural Networks

This paper presents the design of a neural network to determine the categories of Sleep Apnea (SA) or apneic events using Composite Multiscale Sample Entropy (CMSE) as a feature extraction technique. The designed neural network has the ability to process and classify apneic events, maintaining the accuracy levels of apnea scoring of laboratory polysomnography (PSG) which remains the gold standard of sleep monitoring and scoring to date. Additionally, this paper shows the extent to which each individual signal monitored in polysomnography has the ability to independently detect apneic events, which would be useful in the implementation in a portable wearable device.