发表机构
Seoul National University Hospital; Rosota Inc.; Seoul National University; Eulji University College of Medicine（首尔大学医院; Rosota公司; 首尔大学; 乙支大学医学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出从外科医生手持器械记录演示训练双臂手术机器人策略的端到端流程，通过传感器记录器捕获运动，训练扩散策略并在活兔体内成功完成阑尾切除术，验证了无需机器人演示的可行性。
AI 中文摘要
大多数微创手术仍使用手持腹腔镜器械进行，手术结束时外科医生的器械运动学数据会丢失，仅保留内窥镜视频。本文提出了一种端到端流程，可在手术室中捕获该运动，并利用其训练手术机器人策略，并在活体动物上进行了验证。我们引入了一种手术器械状态记录器，它安装在标准腹腔镜器械的杆上，通过惯性传感器、飞行时间传感器和霍尔传感器恢复其姿态和钳口状态，无需外部摄像头或追踪器。数据流程针对机器人地面真值测量每个传感器通道的延迟，并在形成观测-动作对之前对齐各通道。在这些演示中，我们训练了一个具有微调DINOv3骨干的扩散策略，并通过在从离体兔阑尾深度图重建的物理模拟器中进行闭环滚动来选择其设计。随后，该策略在来自四只活兔的849个体内演示上重新训练，并在另外四只装有电外科设备的活兔上部署。在外科医生选择手术阶段的情况下，该策略在四只动物中的三只中完成了阑尾切除术。结果表明，从外科医生自身器械记录的演示足以在体内训练、选择和部署双臂手术策略。机器人仅作为传感器校准的时间参考和执行器，不采集任何演示。两个演示语料库均已发布，以支持未来的手术机器人学习研究。
英文摘要
Most minimally invasive surgery is still performed with hand-held laparoscopic instruments, and the surgeon's instrument kinematics are lost when the operation ends; only the endoscope video is kept. This paper presents an end-to-end pipeline that captures this motion in the operating room and uses it to train a surgical robot policy, validated on live animals. We introduce a surgical instrument-state logger that mounts on the shaft of a standard laparoscopic instrument and recovers its pose and jaw state from an inertial sensor, a time-of-flight sensor and a Hall sensor, with no external camera or tracker. A data pipeline measures the latency of every sensor channel against a robot ground truth and aligns the channels before forming observation-action pairs. On these demonstrations we train a diffusion policy with a fine-tuned DINOv3 backbone, selecting its design by closed-loop rollouts in a physics simulator reconstructed from depth maps of an ex-vivo rabbit appendix. The policy is then retrained on 849 in-vivo demonstrations from four live rabbits and deployed on four additional live rabbits with electrosurgery armed. With the surgeon selecting the surgical phase, the policy completed the appendectomy in three of the four animals. The results show that demonstrations recorded from a surgeon's own instruments are sufficient to train, select and deploy a bimanual surgical policy in vivo. The robot serves only as the timing reference for sensor calibration and as the executor, and collects no demonstrations. Both demonstration corpora are released to support future surgical robot learning research.
CommentsSubmitted to IEEE ICRA 2027
