Development of a low-cost fall intervention system for hospitalized dementia patients

The current state of the art approach to preventing falls of hospitalized elderly adults with dementia is to use a video surveillance setup in each of the hospital rooms and have hospital personnel continuously monitor the video feeds. In this research, we are developing a motion monitoring system to reduce the number of accidental falls among patients at acute risk, while preserving their privacy. The prototypical system includes five accelerometer-based wireless sensors that are placed on the wrists, ankles, and chest of a patient. The system senses the movements and postures of the patient and transmits the information wirelessly to a remote base station. The received motion information is processed in real-time and used to animate a 3D avatar that figuratively represents the movements of the patient. The 3D avatar is intended to give care staff early warning of patient wakefulness, agitation, and of patients attempt to arise from the bed without assistance, while preserving the privacy of the patients. This research also aims to develop predicative algorithms to detect fall-antecedent activity and provide an early warning to care personnel. The base station keeps the captured video and the received motion information synchronized in time and stores them together into a database. The stored video and motion information can be played back in time to plot the motion information as real-time signals on a screen that is synchronized with the captured video of the patient. This paper provides background, motivation, and current state of the art approaches to reducing falls in hospitals. Our prototypical system has undergone preliminary testing successfully on several elderly patients at William Beaumont Hospital at Royal Oak, MI, USA.