Polyp image identification system and method

Abstract

The invention discloses a polyp image identification system. The system comprises an image processor, a video collector and a plurality of program modules, wherein the program modules include an image obtaining module, an image identification module, an algorithm processing module and a prompt processing module; the image obtaining module is used for splitting a video into a plurality of static images frame by frame; the image identification module is used for introducing the static images into a deep convolutional neural network identification engine to obtain a pixel-level probability graph of a plurality of identification targets; the algorithm processing module is used for performing targeted optimization on the input probability graph to remove environmental interference except main target features so as to judge the position of a polyp; and the prompt processing module is used for marking the judged position of the polyp. The invention furthermore discloses a polyp image identification method of the polyp image identification system. The polyp image identification system and method has high sensitivity and high specificity at the same time; the position of the polyp in an endoscopic image can be accurately identified; and the missing identification and false identification rates of polyp detection are remarkably reduced.

Description

technical field [0001] The invention relates to an image recognition system and method, and also relates to a recognition system and method with high sensitivity and high specificity for polyp tissue, belonging to the field of computers. Background technique [0002] Colorectal cancer develops from precancerous lesions such as dead adenomatous polyps and early cancers. It is a malignant tumor with the highest incidence rate at home and abroad, and its prognosis is poor. However, removal of precancerous lesions such as adenomas can effectively avoid the occurrence of interval colorectal cancer and reduce the risk of death from colon cancer; The survival rate can reach more than 90%. Therefore, how to improve the clinical detection rate of colonic adenoma and polypoid lesions of various pathological types is a research direction focused on in cancer medicine research. [0003] Mahmud et al. mentioned in the academic paper "Computer Vision and Augmented Reality in Gastrointes...

AI Technical Summary

Problems solved by technology

[0007] However, what clinical medicine needs is a recognition system with a sensitivity and specificity that can exceed 90% at the same time; lack of clinical value

In the existing technology, when the recognition sensitivity is high, the specificity of CNN is poor, resulting in the misidentification of a large number of normal tissues, such as bulges, obvious blood vessel areas, unreal areas of the image caused by light interference, etc.; and when the specificity is high , the sensitivity of CNN is greatly reduced, and many polyps cannot be effectively identified

[0008] In addition, flat polyps, small polyps, isochromatic polyps, and some lesions with a high probability of transformation into early cancer, because they do not have obvious bulges or spatial geometric features, the existing technology is poor and sensitive in identifying these special polyps. The accuracy is low, there are a lot of missed identification

Among them, some existing technologies pay too much attention to the morphological characteristics of polyps, and the polyps in the screen will change according to various factors such as the shape of the polyp itself and the shooting angle, and cannot be exhaustive; while other existing technologies emphasize the use of large-scale volume It is easy to confuse the morphological characteristics of polyps with the normal morphological characteristics of the digestive tract, such as raised wrinkled walls, etc. This method can neither effectively detect flat polyps, but also increase false positives for normal raised tissues

Images