Comparison of sleep parameter with actigraphy and IR-UWB sensor based on Polysomnography

摘要

The polysomnography(PSG) is a gold standard sleep disorder diagnosis method. However, PSG requires the wearing of many sensors and the trained inspection specialist which lead inspection cost high. In order to overcome this problems, Various sensors such as wrist actigraphy are developed that could resolve complexity and cost of PSG. In this reseach, we deveoloped wake-sleep classification algorithm based on Impulse radio-ultra wide band (IR-UWB) radar and compared with the actigraphy based on the PSG. To identify the proposed algrithm performance, we used 5928 epoch (=30 seconds) samples of sleep scored data obtained from PSG, actigraphy and IR-UWB radar at the same time. From the samples, the sensitivity, specificity and agreement of the proposed algorithm and actigraphy were calculated and compared. In result, Proposed algorithm based on IR-UWB radar showed better specificity and agreement compared to the actigraphy.