A Unified Interaction Control Framework for Safe Robotic Ultrasound Scanning with Human-Intention-Aware Compliance

TL;DR

The paper tackles safe, efficient robotic ultrasound scanning amid frequent human-robot interactions by introducing a unified interaction control framework that discriminates between human-intended and human-unintended actions and adapts compliance accordingly. It leverages a hierarchical, passivity-guaranteed controller with a two-level task decomposition, a five-weight mode-switching scheme, and smooth transitions to couple main and null-space behaviors. The approach is analyzed for stability and passivity and validated in real-world carotid ultrasound experiments, demonstrating reliable performance during intended interventions and resilience to unintended collisions without compromising the primary imaging task. Practically, this framework enables safer, more flexible doctor-in-the-loop ultrasound automation suitable for crowded clinical environments.

Abstract

The ultrasound scanning robot operates in environments where frequent human-robot interactions occur. Most existing control methods for ultrasound scanning address only one specific interaction situation or implement hard switches between controllers for different situations, which compromises both safety and efficiency. In this paper, we propose a unified interaction control framework for ultrasound scanning robots capable of handling all common interactions, distinguishing both human-intended and unintended types, and adapting with appropriate compliance. Specifically, the robot suspends or modulates its ongoing main task if the interaction is intended, e.g., when the doctor grasps the robot to lead the end effector actively. Furthermore, it can identify unintended interactions and avoid potential collision in the null space beforehand. Even if that collision has happened, it can become compliant with the collision in the null space and try to reduce its impact on the main task (where the scan is ongoing) kinematically and dynamically. The multiple situations are integrated into a unified controller with a smooth transition to deal with the interactions by exhibiting human-intention-aware compliance. Experimental results validate the framework's ability to cope with all common interactions including intended intervention and unintended collision in a collaborative carotid artery ultrasound scanning task.

Figures (8)

• Figure 1: An illustration of the human-intention-aware compliance during a carotid artery examination by an ultrasound scanning robot. (a) The doctor grasped the probe to apply coupling gel on the probe, and the robot stopped its current task and followed. (b) The doctor collided with the robot by his leg by accident, and the robot continued its main task with compliance.
• Figure 2: The human-robot interactions considered in this work. The interactions are divided into intended- and unintended- situations and are further classified as several modes based on vision and force information.
• Figure 3: The structure of our proposed method, where multiple modes are integrated into a unified control input with smooth transitions to deal with the interactions by exhibiting human-intention-aware compliance.
• Figure 4: The experimental setup: (a) The hardware system consists of an ultrasound machine, an RGBD camera, a manipulator, an FT sensor, an ultrasound probe, and PCs; (b) An ultrasound image, where the red circle indicates the position of the cross-section carotid artery
• Figure 5: Snapshots of experiment 1: (a) $t=11.5\mathrm{s}$. The probe was approaching the desired scanning area in the Recovery Mode; (b) $t=37\mathrm{s}$. The probe retreated while the patient tilted forward; (c) $t=55\mathrm{s}$. The probe was temporarily removed from the patient's neck; (d) $t=68\mathrm{s}$. The probe followed the desired scanning trajectory while the patient turned to the side; (e) $t=87\mathrm{s}$. The probe was held by the doctor for coupling gel to be applied; (f) $t=130\mathrm{s}$. The probe was operating in the Scanning Mode.