The invention relates to a large intestine endoscope image analysis method based on artificial intelligence. According to the invention, a large intestine endoscope image analysis method based on artificial intelligence,For 8 standard images of large intestine endoscopy, the stages and of multiple image classification models based on artificial intelligence are formed;A plurality of image classification models corresponding to each object part and adjacent parts of the 8 standard images of colorectal endoscopy are formed respectively;Insert the large intestine endoscope probe into the large intestine through the anus, recover the large intestine endoscope probe from the appendix to the anus, and photograph the standard 8 parts of the large intestine endoscope;For the images taken during shooting, multiple image classification models are used to automatically classify and identify the anatomical position of the team leader,Check whether images of standard 8 parts have been taken;The images of 8 standard parts of large intestine endoscopy were classified according to the types of lesions by using the image classification model of tissue biological examination,Find out the lesions and diseases of the input image according to the classification and record the location,Imaging includes the stage of predicting and displaying candidate areas for tissue biopsy.
展开▼
