This project uses the new technology of CT image combined with AI intelligence, and analyzes CT image by artificial intelligence algorithm to judge the incidence and specific location of colorectal polyps, so as to provide an efficient and accurate detection method for early colorectal cancer and reduce the workload of doctors. The core of this project is the artificial intelligence algorithm, we tried the traditional target detection route and image segmentation route. Finally, the U-Net in the image segmentation route is selected as the algorithm model, and the following improvements are made to the algorithm: the backbone feature extraction network is deepened, the ability of the network to extract features is strengthened, and the weighted cross entropy is designed as the loss function. Finally, the detection accuracy rate of 87% was achieved, which met the detection requirements. At the same time, we have developed a matching doctor-side and patient-side small program. This set of platforms provides the following usage scenarios: first, patients do CT detection in hospitals or medical places, and the CT images generated by the detection are transmitted from the hospital intranet to our algorithm model for detection, and the detection results will be sent to the doctor-side display and the patient-side medical applet simultaneously. After reviewing the test results, the doctor will give medical advice and then send it to the patient's medical applet for the patient to view. It can not only improve the efficiency and accuracy of colorectal polyps detection, but also make it more convenient for patients to query results.