Autonomous navigation challenged for flexible robotic colonoscopes
ARCGym: Benchmarking Deep Reinforcement Learning in Autonomous Robotic Colonoscopy
Robotics
Summary
Navigating inside the colon with robotic tools is tricky because the colon is long, flexible, and the robots must carefully move without causing harm. The authors created ARCGym, a computer simulation that mimics realistic colon shapes and robotic colonoscopes to help improve autonomous control using images. They tested different robot types and tasks, and found that guiding flexible colonoscopes using robotic actions, especially those mimicking how doctors insert scopes, is still difficult. Their simulation environment aims to help improve learning methods for safe and effective robotic colonoscopy.
What this means in practice
- •For robotic medical device developers: Test robotic colonoscope navigation algorithms in realistic colon models before clinical trials.
- •For medical simulation trainers: Use ARCGym for training autonomous control of flexible endoscopes in simulated clinical scenarios.
Authors
Guanglin Ji, Martina Finocchiaro, Kenny Erleben, Hang Yin
Abstract
Simulations for learning-based autonomous colonoscopic navigation focus mainly on fully actuated capsule robots, failing to capture the contact-rich navigation of long and flexible clinical colonoscopes. We present the Autonomous Robotic Colonoscopy Gym (ARCGym), an open-source reinforcement learning environment and benchmark for image-based navigation in clinically derived deformable colon anatomies. ARCGym supports multiple types of colonoscope robots, spanning capsule robots and flexible endoscopes, with this work focusing on flexible endoscopes including magnetic-driven tip actuation and clinically used proximally translational actuation. This work includes five CT-reconstructed colons representing typical clinical scenarios, a set of clinically meaningful navigation subtasks, and unified success metrics. We introduce a reward combining depth-based lumen alignment with a lumen-visibility score to improve learning under occlusions. Experiments across tasks, robots, and anatomies show that autonomous navigation remains challenging for both magnetic-driven and proximal-insertion flexible robots, with proximal-insertion actuation remaining an open problem.