CP-CHILD.zip (186.7 MB)
CP-CHILD.zip
CP-CHILD
records the colonoscopy data of children. It contains two datasets CP-CHILD-A and CP-CHILD-B. The CP-CHILD-A dataset contains 8000 RGB images taken by Olympus PCF-H290DI, including 1000 colonic polyp images and 7000 normal or other pathological images. The CP-CHILD-B dataset contains 1500 RGB images taken by FUJIFLIM EC-530wm, including 400 colonic polyp images and 1100 normal or other pathological images. Using the deep learning model on this dataset, we can classify the objects in the dataset. Neural network's self-learning and self-training ability can improve the adaptability, efficiency and detection accuracy of children's colonoscopy CAD.