VISUAL AIDED POSITIONING SYSTEM DESIGN FOR ULTRASONIC SCANNING BASED ON AZURE KINECT

Although ultrasound is a convenient, safe, and non-invasive diagnostic method in clinical practice, due to the significant differences in the level of diagnosis between urban and rural areas, developed areas and remote areas in China, there is also a large demand for ultrasound examinations and there is a huge conflict between the shortage of experienced sonographers and the uneven distribution of high-quality medical resources. Aimed at the above situation, a visual aided system for robotic arm ultrasound scanning using Azure Kinect as a sensor is proposed and designed to guide the robotic arm for automated ultrasound scanning in this study. The system used Azure Kinect to collect RGBD images of patients in real time, locate key points on the human body, and guide the operation of ultrasound scans on the human’s neck. The experimental results show that under the guidance of the Azure Kinect visual positioning system, autonomous ultrasound scanning can be realized, which shortens the routine scanning time, improves the efficiency of diagnosis and treatment for hospital, and saves the time and effort of doctors.