Lesion data collection method, lesion identification method and computer device

Abstract

The invention relates to a lesion data collection method, a lesion identification method and a computer device. The lesion data collection method includes methods that a lesion image sample of a to-be-detected part is obtained; according to a standard pathological report, the lesion position in the lesion image sample is found; according to lesion characteristics corresponding to the lesion position, the lesion part under an electronic endoscope is back-tracked. The lesion identification method is further disclosed, the lesion identification method comprises steps that lesion data is collected, the artificial intelligence deep learning network is utilized for training according to the lesion data to obtain a standard lesion identification model; the model is utilized to identify lesion data of a new to-be-detected part, and the lesion condition is determined. The computer device is further disclosed and includes a processor, a memory and a computer program which is stored on the memoryand operable on the processor. The lesion data collection method, the lesion identification method and the computer device are advantaged in that the lesion position can be accurately found through the electronic endoscope, and the probability of detecting the lesion in advance can be greatly improved.

Description

technical field [0001] The invention belongs to the field of early cancer detection of digestive tract, and in particular relates to a method for acquiring lesion data, a method for identifying lesion and computer equipment. Background technique [0002] The identification of early cancer is through endoscopic screening, detection of suspicious lesions, and finally pathological identification as the standard. Although the image recognition of digestive diseases such as gastrointestinal endoscopy has made great progress, it still mainly depends on pathological recognition to determine the nature of tumors. Pathological examination is the most standard and reliable method to identify tumors. It is to make a smear of the secretion of the diseased organ, or directly remove a small piece of tissue from the lesion to make a slice or print, and place it under a microscope to observe its cell shape, structure, etc., to determine the nature of the tumor. [0003] However, such iden...

AI Technical Summary

Problems solved by technology

[0003] However, such identification time may be relatively long. If the pathological results can be predicted directly through endoscopic judgment, then the pathological results can be identified in advance, which has far-reaching significance for the diagnosis and treatment of early cancer

However, at present, there is a problem of how to match pathological identification with the position of endoscopic lesions in the process of pathologically guiding follow-up treatment.

Images