Detection of patient's bed statuses in 3D using a Microsoft Kinect

Patients spend the vast majority of their hospital stay in an unmonitored bed where various mobility factors can impact patient safety and quality. Specifically, bed positioning and a patient's related mobility in that bed can have a profound impact on risks such as pneumonias, blood clots, bed ulcers and falls. This issue has been exacerbated as the nurse-per-bed (NPB) ratio has decreased in recent years. To help assess these risks, it is critical to monitor a hospital bed's positional status (BPS). Two bed positional statuses, bed height (BH) and bed chair angle (BCA), are of critical interests for bed monitoring. In this paper, we develop a bed positional status detection system using a single Microsoft Kinect. Experimental results show that we are able to achieve 94.5% and 93.0% overall accuracy of the estimated BCA and BH in a simulated patient's room environment.