Simulink-based sleep apnea screening model for portable diagnosis

Obstructive sleep apnea (OSA) is a common sleeping disorder resulting in the temporary blockage of the upper airways that can lead to increased hypertension and risk to other cardiovascular diseases. Most sleep apnea cases currently remain undiagnosed due to cost and resource limitations of overnight polysomnography at sleep laboratories. In this project, we demonstrate a Simulink-based, modular and reprogrammable hardware system model design for portable home monitoring of sleep apnea syndromes. Diagnosis is achieved using a novel screening algorithm developed in our earlier projects based on wavelet analysis and feed-forward neural networks, which only requires an input of an overnight single lead ECG recording. Preliminary results indicate that the hardware implementation of our Simulink model produces the equivalent output as the novel algorithm. The componentized nature of the hardware system design allows for rapid deployment of progressively improving ECG-based algorithms to diagnose OSA. This would be a step toward addressing the serious public health issue caused by the under-diagnosis of sleep apnea.