Colorectal cancer is a major global health concern, making early detection and prevention essential for reducing its incidence and mortality. Colonoscopy is widely used for colorectal cancer screening, as it enables direct examination of the colon and the detection and removal of precancerous polyps. However, navigating a flexible colonoscope through the complex and deformable geometry of the colon remains technically challenging and requires considerable clinical expertise. Robotic assistance has the potential to improve navigation, reduce the burden on clinicians, and increase the safety and effectiveness of the procedure.
Despite these advantages, robotic colonoscopy still presents several challenges, particularly in achieving reliable and safe navigation through the complex geometry of the colon. Accurate control of the robotic platform is essential to follow the desired path while adapting to changes in the surrounding anatomy. Addressing these navigation and control challenges is therefore important for the development of effective robotic colonoscopy systems.
The aim of this project is to introduce a pipeline for autonomous navigation of a robotic colonoscopy platform while simultaneously mapping the colon in the insertion phase and detecting polyps along the way. The approach consists of three main sub-pipelines. The first estimates the optimal navigation point using a combination of classical and deep learning-based lumen navigation methods, using monocular camera images at the tip of the endoscope.
The second pipeline is a polyp and anomaly detection algorithm that runs concurrently with the lumen navigation pipeline. The main objective of this pipeline is detecting abnormalities and polyps during the insertion phase, providing information about regions that need monitoring during the retraction phase and reducing the risk of missed polyps. The third pipeline uses a SLAM algorithm to map the colon during the insertion phase. This map can then be used to localise the endoscope during retraction and associate detected polyps with their locations inside the colon.
