发表机构
The Chinese University of Hong Kong; Great Bay University; Dongguan Great Bay Institute for Advanced Study; The University of Hong Kong; Southern University of Science and Technology（香港中文大学; 大湾区大学; 东莞大湾区先进研究院; 香港大学; 南方科技大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出任务相关特征动力学保真度（TR-FDF），开发面向TR-FDF的超声模拟器，实现仅仿真训练的机器人超声策略在体模实验中400次零样本部署成功390次，提升零样本仿真到真实迁移性能。
AI 中文摘要
直接基于B型图像的机器人超声策略需要大量交互数据，而真实机器人的数据采集成本高且受安全限制。仿真提供了可扩展的替代方案，但零样本迁移不仅依赖单帧真实感，还取决于模拟观测是否复现探头运动诱导的任务相关特征变化，我们将这种跨域一致性称为任务相关特征动力学保真度（TR-FDF）。在局部正则性假设下，我们的收缩分析表明，TR-FDF失配对探头运动的敏感性越高，有效闭环收缩裕度越小；而与运动无关的误差主要增大残差误差界。受此分析指导，我们开发了面向TR-FDF的超声模拟器，其结合了共享结构中间域、轨迹级固定噪声和少步条件流生成。在体模实验中，仅在仿真中训练的策略在四个目标平面的400次零样本部署中成功了390次。该模拟器的FID为29.66，生成观测的速度为67.1 Hz。受控干预、消融实验和基线对比显示，TR-FDF敏感性在预测零样本迁移性能方面补充了单帧真实感。
英文摘要
Robotic ultrasound policies operating directly on B-mode images require extensive interaction data, whereas real-robot data collection is costly and safety-constrained. Simulation provides a scalable alternative, but zero-shot transfer depends not only on single-frame realism but also on whether simulated observations reproduce task-relevant feature changes induced by probe motion. We term this cross-domain consistency task-relevant feature-dynamics fidelity (TR-FDF). Under local regularity assumptions, our contraction analysis shows that greater sensitivity of TR-FDF mismatch to probe motion reduces the effective closed-loop contraction margin, whereas motion-independent errors primarily enlarge the residual error bound. Guided by this analysis, we develop a TR-FDF-oriented ultrasound simulator that combines a shared structural intermediate domain, trajectory-level fixed noise, and few-step conditional flow generation. In phantom experiments, a policy trained exclusively in simulation succeeded in 390 of 400 zero-shot deployments across four target planes. The simulator achieved an FID of 29.66 and generated observations at 67.1 Hz. Controlled interventions, ablations, and baseline comparisons showed that TR-FDF sensitivity complements single-frame realism in predicting zero-shot transfer performance.
