RT Journal Article SR Electronic T1 A Probabilistic Approach to the Detection and Diagnosis of Obstructive Sleep Apnea Using Chest Movement Analysis, Recursive Noise Reduction and Machine Learning Algorithms JF bioRxiv FD Cold Spring Harbor Laboratory SP 080515 DO 10.1101/080515 A1 Anshul Tripathi A1 Raj Ramnani YR 2016 UL http://biorxiv.org/content/early/2016/10/18/080515.abstract AB In a novel approach to diagnose Obstructive Sleep Apnea, electronic components, such as an Arduino Mega, a Bluetooth Transceiver, an accelerometer, and a air-aulity sensor, were put together to create a wearable that would detect the frequency of apnea events, detect and diagnose the disorder, and sound an alarm when necessary. A primary consideration was to make the mechanism accessible and affordable, and in doing so, lower the cases of Obstructive Sleep Apnea that go undiagnosed due to the cost and inconvenience associated with the traditional diagnosis method—a Polysomnogram. Bluetooth capability was an additional consideration so that the device would transmit data directly to an android smartphone, eliminating the need for an additional output mechanism. The total cost of the device, quite surprisingly, did not exceed $30, and therein rendered the device an accessible, affordable mechanism for diagnosis. Tests of the device on diagnosed patients yielded data consistent with the diagnosis, with a few false positives as a result of the excessive sensitivity of the sensors.