ArthroCut: Autonomous Policy Learning for Robotic Bone Resection in Knee Arthroplasty

Xu Lu; Yiling Zhang; Wenquan Cheng; Longfei Ma; Fang Chen; Hongen Liao

cs.ROMar 4, 2026

ArthroCut: Autonomous Policy Learning for Robotic Bone Resection in Knee Arthroplasty

Authors: Xu Lu, Yiling Zhang, Wenquan Cheng, Longfei Ma, Fang Chen, Hongen Liao

Organizations: School of Biomedical Engineering, Tsinghua University, Beijing 100084, China. · Longwood Valley MedTech, Beijing 100176, China. · School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

Abstract

Despite rapid commercialization of surgical robots, their autonomy and real-time decision-making remain limited in practice. To address this gap, we propose ArthroCut, an autonomous policy learning framework that upgrades knee arthroplasty robots from assistive execution to context-aware action generation. ArthroCut fine-tunes a Qwen--VL backbone on a self-built, time-synchronized multimodal dataset from 21 complete cases (23,205 RGB--D pairs), integrating preoperative CT/MR, intraoperative NDI tracking of bones and end effector, RGB--D surgical video, robot state, and textual intent. The method operates on two complementary token families -- Preoperative Imaging Tokens (PIT) to encode patient-specific anatomy and planned resection planes, and Time-Aligned Surgical Tokens (TAST) to fuse real-time visual, geometric, and kinematic evidence -- and emits an interpretable action grammar under grammar/safety-constrained decoding. In bench-top experiments on a knee prosthesis across seven trials, ArthroCut achieves an average success rate of 86% over the six standard resections, significantly outperforming strong baselines trained under the same protocol. Ablations show that TAST is the principal driver of reliability while PIT provides essential anatomical grounding, and their combination yields the most stable multi-plane execution. These results indicate that aligning preoperative geometry with time-aligned intraoperative perception and translating that alignment into tokenized, constrained actions is an effective path toward robust, interpretable autonomy in orthopedic robotic surgery.

Figures & tables

Fig. 1: Comparison between vanilla VLA, commercial orthopedic robots, and the proposed ArthroCut framework. Conventional VLA models (top) directly map task inputs to robotic actions but fail to incorporate detailed surgical scene understanding, limiting their robustness in complex environments. Commercial orthopedic robots (middle) use intraoperative data to follow pre-templated paths but lack real-time adaptation and personalization [ 1 ] . In contrast, ArthroCut (bottom) integrates preoperative imaging with temporally aligned multimodal intraoperative data, enabling accurate, real-time decision-making that improves both surgical precision and personalization.

Fig. 2: The overview of our proposed method for robotic bone resection in knee arthroplasty. Preoperative imaging and intraoperative multimodal streams are encoded into tokens and processed by a transformer-based model to generate surgical actions ( $\texttt{<MOVE>},\texttt{<ALIGN>},\texttt{<CUT>}$ ) for autonomous femoral cutting.

Fig. 3: Example of task execution for femoral cutting across five planes using ArthroCut. For each task: Left—text instruction with the initial state (top-right inset of the second column shows the pre-cut bone surface); Middle—generated intermediate goal states; Right—final state upon task completion. Full execution trajectories are provided in the supplementary video.

Method	Metrics
	Success Rate (SR) $\uparrow$
	anterior chamfer	distal femur	anterior condyle	posterior condyle	tibial	posterior chamfer
OpenVLA (ft) [ 17 ]	0.29 (2/7)	0.29 (2/7)	0.29 (2/7)	0.00 (0/7)	0.29 (2/7)	0.00 (0/7)
Diffusion Policy [ 34 ]	0.47 (4/7)	0.29 (2/7)	0.47 (4/7)	0.29 (2/7)	0.00 (0/7)	0.00 (0/7)
RT-2 (ft) [ 7 ]	0.29 (2/7)	0.29 (2/7)	0.29 (2/7)	0.00 (0/7)	0.00 (0/7)	0.00 (0/7)
Octo (ft) [ 35 ]	0.29 (2/7)	0.00 (0/7)	0.00 (0/7)	0.29 (2/7)	0.00 (0/7)	0.00 (0/7)

TABLE I: Bench evaluation on a knee prosthesis (SR). Success threshold

\delta{=}1.5

mm mean surface deviation across the six planes. All methods use the same action grammar and constrained decoding. ft: fine-tune.

Method	Metrics
	Success weighted by Path Length (SPL) $\uparrow$
	anterior chamfer	distal femur	anterior condyle	posterior condyle	tibial	posterior chamfer
OpenVLA (ft) [ 17 ]	0.22 $\pm$ 0.02	0.24 $\pm$ 0.02	0.24 $\pm$ 0.02	0.00 $\pm$ 0.00	0.20 $\pm$ 0.04	0.00 $\pm$ 0.00
Diffusion Policy [ 34 ]	0.42 $\pm$ 0.05	0.23 $\pm$ 0.03	0.43 $\pm$ 0.04	0.22 $\pm$ 0.03	0.00 $\pm$ 0.00	0.00 $\pm$ 0.00
RT-2 (ft) [ 7 ]	0.21 $\pm$ 0.02	0.18 $\pm$ 0.03	0.20 $\pm$ 0.03	0.00 $\pm$ 0.00	0.00 $\pm$ 0.00	0.00 $\pm$ 0.00
Octo (ft) [ 35 ]	0.22 $\pm$ 0.03	0.00 $\pm$ 0.00	0.00 $\pm$ 0.00	0.22 $\pm$ 0.02	0.00 $\pm$ 0.00	0.00 $\pm$ 0.00

TABLE II: Bench evaluation on a knee prosthesis (SPL). Success threshold

\delta{=}1.5

mm mean surface deviation across the six planes. SPL is computed with the protocol and reported as mean

\pm

\sim

SD over episodes. All methods use the same action grammar and constrained decoding. ft: fine-tune.

Fig. 4: Schematic diagram of knee replacement verification after osteotomy. (a) Femoral resections provide a visually flat surface for component seating; (b) the joint shown in full extension after component placement.

Method	Inference time (ms)	Robot exec. time (s)
Vanilla Model	267.14	121.89
ArthroCut (Ours)	302.52	107.33

TABLE III: Comparison of inference latency and robot execution time (mean) between the Vanilla model and the proposed ArthroCut framework.

Method	Ablation		SR $\uparrow$	SPL $\uparrow$
Method	PIT	TAST	SR $\uparrow$	SPL $\uparrow$
Vanilla Model			0.05	0.04
Vanilla Model	✓		0.12	0.10
Vanilla Model		✓	0.68	0.61
ArthroCut (Ours)	✓	✓	0.86	0.75

TABLE IV: Ablation study on the proposed ArthroCut framework. PIT denotes the proposed preoperative imaging tokens and TAST denotes the time-aligned surgical tokens.

Surgical Robotics

Surgical robotics develops robotic tools and systems for minimally invasive, image-guided, and microsurgical procedures.

See all 187 papers

Aug 17, 2026

SurgVIL: Scaling Surgical Robot Imitation Learning with Open-source Surgical Videos

Learning-based surgical robot autonomy requires large-scale demonstrations with synchronized videos and robot actions, but such data are exceedingly rare in clinical or realistic tissue settings because robot kinematics are typically inaccessible outside controlled research systems. In contrast, phantom data collected on research platforms provide accurate action labels but lack the visual diversity of real tissue. We propose SurgVIL, a framework for scaling surgical robot imitation learning using open-source surgical videos. SurgVIL combines kinematically labeled phantom robot demonstrations with surgical videos from open-source datasets and online sources for policy learning. Since these videos lack robot motion labels, we estimate approximate kinematics as weak supervision. We evaluate SurgVIL on two da Vinci robot tasks: needle pick-up and cholecystectomy cutting. Across ACT,

π_0

, and GR00T-H backbones, adding surgical videos substantially improves generalization to real-tissue and out-of-distribution settings, suggesting a scalable path from phantom training toward generalizable surgical robot policies.

Xinhao Chen, JuoTung Chen, Nigel Nelson +4

Laboratory for Computational Sensing and Robotics (LCSR) Johns Hopkins University, USA · Nvidia, USA

Robot Learning from Video Robot Imitation Learning Surgical Robotics Robot Learning Generalization in Robotic Manipulation

Ask PDF

Jul 23, 2026

RL-MACRO: A Cybernetic Closed-Loop Intelligence Framework for Multimodal Adaptive Robotic Craniotomy

Autonomous robotic craniotomy requires continuous regulation of tool-tissue interactions to mitigate mechanical overload and thermal damage while maintaining surgical efficiency. However, this process is inherently partially observable due to unknown, time-varying tissue properties and the inability to directly measure cutting temperatures under physical occlusion. To address these challenges, we propose RL-MACRO, a cybernetic closed-loop intelligence framework that couples multimodal perception, adaptive decision-making, and robotic execution. This framework empowers the surgical robot to autonomously perceive inaccessible states from partial sensory feedback and dynamically optimize its behaviors under uncertain environment. A CNN-LSTM observer first fuses force and sound feedback to reconstruct the hidden temperature state (R^2=0.939, MAE = 1.717 deg C). This reconstructed temperature, alongside multi-sensor features, forms the belief state for an offline Implicit Q-Learning (IQL) policy. A novel dual-head Actor dynamically coordinates the feed rate, spindle speed, and cutting depth to optimize efficiency within strict safety bounds. These decisions are seamlessly translated into spatial motions via online trajectory re-planning and velocity servoing. Experiments on bovine ribs and six ex vivo goat skulls validate the system's robust perception, adaptive recovery from force/temperature excursions, and smooth execution on irregular surfaces, establishing a data-driven cybernetic paradigm for safe and efficient autonomous bone cutting.

Xiao Zhang, Jiaxuan Li, Renzhen Le +9

Dalian University of Technology, Dalian 116024, China · Second Hospital of Dalian Medical University, Dalian 116024, China · Department of Mechanical Engineering, The University of Tokyo, Tokyo 113-8656, Japan

Robotic Control Surgical Robotics

Ask PDF

Mar 9, 2026

An Open-Source Robotics Research Platform for Autonomous Laparoscopic Surgery

Autonomous robot-assisted surgery demands reliable, high-precision platforms that strictly adhere to the safety and kinematic constraints of minimally invasive procedures. Existing research platforms, primarily based on the da Vinci Research Kit, suffer from cable-driven mechanical limitations that degrade state-space consistency and hinder the downstream training of reliable autonomous policies. We present an open-source, robot-agnostic Remote Center of Motion (RCM) controller based on a closed-form analytical velocity solver that enforces the trocar constraint deterministically without iterative optimization. The controller operates in Cartesian space, enabling any industrial manipulator to function as a surgical robot. We provide implementations for the UR5e and Franka Emika Panda manipulators, and integrate stereoscopic 3D perception. We integrate the robot control into a full-stack ROS-based surgical robotics platform supporting teleoperation, demonstration recording, and deployment of learned policies via a decoupled server-client architecture. We validate the system on a bowel grasping and retraction task across phantom, ex vivo, and in vivo porcine laparoscopic procedures. RCM deviations remain sub-millimeter across all conditions, and trajectory smoothness metrics (SPARC, LDLJ) are comparable to expert demonstrations from the JIGSAWS benchmark recorded on the da Vinci system. These results demonstrate that the platform provides the precision and robustness required for teleoperation, data collection and autonomous policy deployment in realistic surgical scenarios.

Ariel Rodriguez, Lorenzo Mazza, Martin Lelis +4

Department of Translational Surgical Oncology, NCT/UCC Dresden, a partnership between DKFZ, MF UKD, TUD, HZDR Dresden, Germany · Cluster of Excellence-CeTI, TUD, Germany · Department of Visceral, Thoracic and Vascular Surgery, Faculty of Medicine and University Hospital Carl Gustav Carus, TUD, Germany

Laparoscopic Surgery Surgical Robotics Robotic Control Healthcare Robotics

Ask PDF

Abstract

Surgical Robotics

Explore similar work

SurgVIL: Scaling Surgical Robot Imitation Learning with Open-source Surgical Videos

RL-MACRO: A Cybernetic Closed-Loop Intelligence Framework for Multimodal Adaptive Robotic Craniotomy

An Open-Source Robotics Research Platform for Autonomous Laparoscopic Surgery