Autonomous colon navigation system improves endoscopic robot control
ColoACT: Multi-Cue Action Chunking for Smooth Autonomous Colon Navigation on a Self-Propelled Endoscopic Robot
Robotics
Summary
Colonoscopies can be hard to perform automatically because the colon is soft and twists, and the camera images inside can be hard to interpret due to shiny surfaces and few unique features. The authors designed a robot with a new navigation system that uses color images combined with estimated depth and shape cues to recognize important patterns in the colon lining. Their system can smoothly control the robot through different colon segments, including turns and bends, tested in pig colons outside the body. This work shows progress toward safer and less labor-intensive colonoscopy procedures.
What this means in practice
- •For medical device developers: Develop autonomous colonoscopy robots capable of smoother and more reliable navigation through complex colon anatomy.$Commercial implications: Enables building endoscopic robots that reduce operator workload and risks during colonoscopy procedures, potentially commercialized as medical devices.
- •For robotic surgery engineers: Integrate multi-cue visual and control strategies to improve autonomous navigation in deformable, low-feature environments like internal anatomy.
Authors
Jian Hu, Shujing He, Leixin Chang, Zongze Li, Ding Huang, Chaoyang Shi, Chengzhi Hu
Abstract
Autonomous colonoscopic navigation can reduce operator burden and the risk of loop formation or tissue trauma, but remains challenging due to deformable anatomy, weak-texture and specular endoscopic visuals, and contact-rich viscoelastic interactions. Existing methods either rely on geometry-driven pipelines, which are efficient and interpretable yet brittle due to manually engineered features and switching logic, or adopt learning-based policies, whose inferred depth/geometry can become temporally inconsistent or overly smooth under weak texture and specular highlights while simulation-trained variants (e.g., deep reinforcement learning) may further suffer from a sim-to-real gap. We propose ColoACT, an autonomous navigation system that integrates an RGB-D-E based Action Chunking Transformer policy (ColoACT policy) for a compact self-propelled Bevel-Gear-Based Endoscopic Robot (BGER). The ColoACT policy augments RGB with estimated relative depth and a gradient-based pseudo-elevation map to enhance fold-ridge saliency and other high-frequency geometric cues, and enables smooth continuous control of the BGER by predicting overlapping action chunks and fusing them via temporal ensembling. In different \textit{ex-vivo} porcine colons (approximately 60 cm), our system achieves success rates of 85.4\% and 72.5\% in straight and curved segments, respectively, and achieves 70\% success in 90-degree turns and 60\% in double-bend sequences, with feasibility further demonstrated in challenging triple-bend segments. The project page is available at: https://Adamhu1.github.io/ColoACT/.