Imitation Learning for Robot Assistance in Open Surgery: A Multi-Policy Evaluation on Suture Following

2026-05-27

AI Summary

This study presents the first systematic evaluation of general-purpose imitation learning for collaborative open surgical suturing, aiming to automate robotic assistance in the "grasp-pull-release" maneuver. Leveraging 160 teleoperated demonstration trajectories, the authors benchmark four state-of-the-art policies—ACT, Diffusion Policy, SmolVLA, and π₀—on an open-source robotic platform across varying viewpoints, background conditions, and dataset sizes. The results demonstrate that π₀, when integrated with a pretrained vision-language backbone, achieves superior performance in data efficiency, background robustness, and trajectory smoothness. Under ideal conditions, it attains a task success rate of 50–75%, and in real-world suturing trials, it achieves a 92% completion rate, with primary failures attributed to depth perception errors.

Abstract

This study presents the first evaluation of general-purpose imitation learning for surgeon-robot collaborative assistance in open surgery, targeting suture following: the grab-pull-release motion an assistant performs at every stitch. We collect 160 teleoperated demonstrations (32,374 frames) on an open-source robot arm, benchmark four architecturally diverse imitation learning policies (ACT, Diffusion Policy, SmolVLA, $π_0$) across 28 trained models evaluated in 32 configurations along three clinically motivated dimensions: dataset size, camera viewpoint, and background variation. Our results demonstrate that under ideal conditions, the four policies achieve $50$-$75\%$ task success, with depth error as the dominant failure mode across all architectures. Among all policies, $π_0$ achieves the strongest results with a pretrained vision-language backbone, demonstrating superior data efficiency, greater robustness to background variation, and smoother trajectories compatible with surgical workflow. When deployed in a surgeon-robot suturing trial, $π_0$ yields a $92\%$ stitch completion rate. These findings establish collaborative robotic assistance in open surgery as a feasible target for imitation learning and highlight depth perception and end-effector design as key priorities for clinical translation.

Problem

• Imitation Learning
• Robot Assistance
• Open Surgery
• Suture Following
• Collaborative Robotics

Innovation

• Imitation Learning
• Surgical Robotics
• Suture Following
• Vision-Language Model
• Multi-Policy Benchmarking