SonoGym: High Performance Simulation for Challenging Surgical Tasks with Robotic Ultrasound

TL;DR

SonoGym is a scalable, realistic simulation platform for robotic ultrasound tasks that enables training advanced AI agents for complex surgical procedures, addressing a key gap in simulation environments for medical robotics.

Contribution

Introduces SonoGym, a high-performance simulation environment supporting realistic, real-time ultrasound data and enabling training of DRL and IL agents for robotic surgery tasks.

Findings

Successful training of AI agents for anatomy reconstruction

Demonstrated policy learning across multiple surgical scenarios

Highlighted current limitations in clinical environments

Abstract

Ultrasound (US) is a widely used medical imaging modality due to its real-time capabilities, non-invasive nature, and cost-effectiveness. Robotic ultrasound can further enhance its utility by reducing operator dependence and improving access to complex anatomical regions. For this, while deep reinforcement learning (DRL) and imitation learning (IL) have shown potential for autonomous navigation, their use in complex surgical tasks such as anatomy reconstruction and surgical guidance remains limited -- largely due to the lack of realistic and efficient simulation environments tailored to these tasks. We introduce SonoGym, a scalable simulation platform for complex robotic ultrasound tasks that enables parallel simulation across tens to hundreds of environments. Our framework supports realistic and real-time simulation of US data from CT-derived 3D models of the anatomy through both a physics-based and a generative modeling approach. Sonogym enables the training of DRL and recent IL agents (vision transformers and diffusion policies) for relevant tasks in robotic orthopedic surgery by integrating common robotic platforms and orthopedic end effectors. We further incorporate submodular DRL -- a recent method that handles history-dependent rewards -- for anatomy reconstruction and safe reinforcement learning for surgery. Our results demonstrate successful policy learning across a range of scenarios, while also highlighting the limitations of current methods in clinically relevant environments. We believe our simulation can facilitate research in robot learning approaches for such challenging robotic surgery applications. Dataset, codes, and videos are publicly available at https://sonogym.github.io/.