Capturing nursing interactions from mobile sensor data and in-room sensors

Abstract

In this paper, we show two approaches for capturing nursing interactions in a hospital: 1) finding nursing intervals from mobile sensors with accelerometers and audio on nurses, and 2) recognizing nurses' entrance to a patient's room from in-room sensors of bed, loudness, and illuminance sensors. For 1), we firstly detect the nurses' entrance to the patient's room by walking detection from accelerometers and noise level on mobile sensors, and detect the interval of interaction between nurses and the patient. For 2), we recognize the nurse's entrance to the patient's room with in-room sensors, using separate algorithms between day and night. We developed the algorithms using the sensor data collected in a cardiovascular center in a real hospital for one year. It could be a important baseline technique to find valuable intervals from long and big data of sensors. © 2013 Springer-Verlag Berlin Heidelberg.