Imitation Learning of Compression Pattern in Robotic-Assisted Ultrasound Examination Using Kernelized Movement Primitives

Abstract

Vascular diseases are commonly diagnosed using Ultrasound (US) imaging, which can be inconsistent due to its high dependence on the operator’s skill. Among these, Deep Vein Thrombosis (DVT) is a common yet potentially fatal condition, often leading to critical complications like pulmonary embolism. Robotic US Systems (RUSs) aim to improve diagnostic test consistency but face challenges with the complex scanning pattern requiring precise control over US probe pressure, such as the one needed for indirectly detecting occlusions during DVT assessment. This work introduces an imitation learning method based on Kernelized Movement Primitives (KMP) to standardize the contact force profile during US exams by training a robotic controller using sonographer demonstrations. A new recording device design enhances demonstration acquisition, integrating with US probes and enabling seamless force and position data recording. KMPs are used to link scan trajectory and interaction force, enabling generalization beyond the demonstrations. Our approach, evaluated on synthetic models and volunteers, shows that the KMP-based RUS can replicate an expert’s force control and US image quality, even under conditions requiring compression during scanning. It outperforms previous methods using manually defined force profiles, improving exam standardization and reducing reliance on specialized sonographers.