Design and Application of an Intestinal Multi-Robotwith YOLOv8-DCP Algorithm for Accurate PolypDetection

This paper presents a novel multifunctional intestinal robot that integrates capabilities for dual biopsy and drug release with animproved YOLOv8-DCP algorithm for enhanced polyp detection under endoscopic conditions. The proposed system addressesthe limitations of single-function robots by enabling multiple medical operations within a single treatment session. The YOLOv8-DCP algorithm, designed specifically for the complex intestinal environment, incorporates a convolution and attention fusionmodule (CAFM), a Dense Convolutional Feature Extraction Module (DC2F), and a P6 detection head. Experimental resultsdemonstrate significant improvements in detection accuracy, with the YOLOv8-DCP model outperforming basic models acrossall metrics, particularly in the detection of large polyps. This study offers a promising approach for the more accurate andefficient detection and treatment of intestinal polyps, contributing to the advancement of clinical applications in intestinalrobotics.