This paper presents a fully autonomous robotic ultrasound system (FARUS) for thyroid scanning and malignant nodule identification. The system uses human skeleton point recognition, reinforcement learning, force feedback, and Bayesian optimization to overcome challenges in locating thyroid targets and optimizing probe orientation. Experimental results show that FARUS produces high-quality ultrasound scans comparable to those from clinicians and has the potential for thyroid nodule detection and ACR TI-RADS calculation.
Publisher
nature communications
Published On
May 11, 2024
Authors
Kang Su, Jingwei Liu, Xiaoqi Ren, Yingxiang Huo, Guanglong Du, Wei Zhao, Xueqian Wang, Bin Liang, Di Li, Peter Xiaoping Liu
Tags
robotic ultrasound
thyroid scanning
malignant nodule identification
reinforcement learning
Bayesian optimization
automated healthcare
Related Publications
Explore these studies to deepen your understanding of the subject.