Development of wearable sensor system and deep network architectureto provide sleeping position detection to prevent pressure injurywearable inertial sensors and deep network architecture

Pressure ulcers are one of the serious health issues that arise due to impaired blood circulation in individuals who are required to remain in a lying or sitting position for prolonged periods. The aging population and sedentary lifestyles have increased the prevalence of these wounds. In this study, a system combining wearable sensor technologies and deep learning algorithms was developed to monitor lying positions and provide alerts. The system tracks positions through three wireless IMU sensors (inertial measurement units) placed on the chest and legs of the patient. Data from the sensors is processed using data fusion algorithms that ensure high accuracy in determining position and orientation angles. Machine learning and deep learning-based algorithms developed in this study demonstrated that lying positions can be detected with an accuracy of 88%-99%. The developed system offers an innovative approach to improving personalized healthcare services for the management of pressure ulcers and reducing the workload of healthcare professionals. Keywords: Pressure injury prevention, Lying position detection, Inertial sensor data, Deep learning, Data classification