AI-enabled wearable sensing system to perform conformity test in healthcare

There has been tremendous research efforts aimed at improving lower back healthcare through wearable sensing systems, yet the potential for improvement still lacks due to conformity test necessities in the process. This project demonstrates the development of an AI-based wearable sensing system to perform conformity tests in the healthcare domain. The project focuses on the lower back or lumbar-pelvic movement monitoring to tackle back pain problems. The sensing system is divided into three different simultaneously operating units; a wearable watch, a lower back device, and a computer vision unit. An IoT network is created between the devices to measure real-time dual IMU sensor data and perform machine learning to generate an appropriate posture control signal. The computer vision model is trained with the relevant data for different rehabilitation exercises for patients using OpenCV in Python. Preliminary results show improvements with higher accuracies up to 90% when compared to other posture control methods.