Intelligent Robotic Ultrasound Imaging: Collaborative Sensing and Active Perception
MCML Authors
Abstract
Abstract
This thesis aims to reduce the reliance on human expertise in maneuvering the ultrasound (US) probe and interpreting the acquired US images. It presents a low-cost 3D US system, a human-robot collaborative scanning framework, and actively incorporates temporal cues into deep learning-based perception models. By integrating and emulating human expertise with robotic precision and the temporal dynamics of US images, it enables safer, more intelligent US procedures with reduced clinician fatigue and paves the way for autonomous US examination systems.
BibTeXKey: Hua25