Detection of Falling Behavior of Patient from Bed Using Three-Dimensional Recognition

Abstract

WHO reports that the accident that a patient falls off the bed is the largest incident occurring in the hospital. Each medical institution is making efforts to eliminate the fall, but the incident has not decreased. We in clinical research constantly measure and record the behavior of the patient on the bed like a vital sensor, and we detected actions that lead to falling from various patient behaviors in three dimensions. We were able to detect signs of fall by constantly measuring patient behavior. In addition, by visualizing patient behavior, we caught quantitatively the changes in behavior of patients who were known only by the sense of nurses. From these, it is possible to prevent the patient from falling down before it falls from the bed, and it can show changes in patient behavior in the form of graphs, as one of the data for evidence-based medical care in nursing.