APNIWAVE: An Efficient Radar-Based Sleep-Apnea Screening Device for Use at Home

Obstructive Sleep Apnea and Hypopnea Syndrome (OSAHS) is a widespread non-communicable disease in which the vast majority of cases remains undetected. To solve this problem, we introduce APNIWAVE which includes an ultra-wideband radar sensor, a Raspberry Pi device as well as Machine Learning algorithms to detect OSAHS by observing the breathing pattern of the screening patient. The developed solution is based on low-cost components, it is very convenient to use and can be particularly used for screening patients at their home, adding in this way zero additional burden to the overloaded third-grade units worldwide. The proposed solution was applied to 11 patients with OSAHS. The data were collected during their entire sleep interval (ranging between 5.5 - 8 hours) and used for the training of five Machine Learning algorithms. The highest classification accuracy was 88% and was achieved using random forest, validating the efficacy of the developed framework.