Enabling Guardian Angels: Designing and Constructing a Wireless Nurse Call System with IMU-Based Fall Detection for Enhanced Patient Safety

Abstract

Falling poses a significant health concern across all age groups, with particular severity among the elderly. Hospitalized patients, in particular, are vulnerable to injuries and even death due to falls. While patient supervision is essential for fall prevention, constant proximity between patients and healthcare staff is not always feasible. To tackle this challenge, this study aimed to develop a solution that enables immediate assistance for patients who are distant from the nurse call button when a fall occurs. The study employed the IMU sensor, which combines an accelerometer and a gyroscope. This sensor served as a transmitter to detect gravity acceleration and magnitude when a fall event takes place. The data obtained from the IMU sensor were further processed using an Arduino Uno microcontroller. The sensor was integrated into a belt worn around the waist of the participants, who performed various movements such as falling facing down, falling up, falling to the right, falling to the left, standing then sitting, and sitting then standing. The experimental tests yielded compelling results, with all trials achieving an accuracy rate of 81.7%. The accuracy was determined by analyzing the confusion matrix, which enabled accurate calculations. The utilization of this innovative tool significantly reduces the risk of patients experiencing detrimental outcomes following falls by promptly notifying medical personnel, even when they are distant from the nurse call button. Moreover, the implementation of this tool enhances overall safety for hospitalized patients, especially those at a high risk of falling. Future research can explore the integration of additional sensors or the development of more sophisticated algorithms to further enhance the accuracy and efficacy of this tool.