Sleep Apnea and Hypopnea Detection Algorithm

Sleep-related breathing disorders such as sleep apneas in most cases are related to cardiological or neural diseases. Moreover, they make an impact on the human health state causing an inefficient sleep, fatigue, exhaustion, drowsiness and low activity level. A traditional study for sleep breathing disorders is polysomnography. It is complex and expensive to significant part of patients with such symptoms and requires to be performed in a special room in the hospital. Holter monitoring systems are widely used in cardiological diagnostics. In some cases they are also used for sleep apneas detection via thoracic excursion signal measured by impedance pneumography. In this article an algorithm for Cheyne-Stokes breathing detection and sleep apnea and hypopnea episodes determination inside one is purposed. It allows to expand abilities of Holter monitors and perform screening of potential patient at home, faster, without additional wires or discomfort. To check this algorithm 19 night records where analyzed, of them 10 to train the model, 9 to test. The determination of apnea type, i.e. central versus obstructive is not the focus of this paper.