Automatic Generation of Autonomous Ultrasound Scanning Trajectory Based on 3-D Point Cloud

Abstract

The fusion of robotics and ultrasound scanning has been used to overcome the shortcomings of traditional ultrasound examinations, such as high variability and low repeatability of ultrasound images acquired by operators. However, there are huge differences in the surface morphology of different human bodies, and the current clinical practice still relies heavily on the manually generating a suitable scanning trajectory for the free-form surface. In this paper, a practical strategy for automatic generation of ultrasound scanning trajectory on breast surfaces is proposed, which can automatically and efficiently generate scanning trajectory for any free-form breast surface. First, the probe attitude and contact model of ultrasound automatic scanning are analyzed. Second, an end-to-end automatic trajectory generation workflow based on 3D point clouds is proposed. Furthermore, a trajectory point offset strategy considering skin deformation and a simultaneous double breast scanning strategy are proposed. In addition, a probe attitude normal admittance controller based on position inner loop control is designed. The experimental results show that the overall time of trajectory generation is less than 4 s. The automatic scanning is 30.84 times more stable in controlling the contact force than the manual scanning, which effectively improves the repeatability of the scanning.