Image recognition system for distinguishing atrophic gastritis and gastric cancer, equipment and medium

Abstract

The invention provides an image recognition system for distinguishing atrophic gastritis and gastric cancer, equipment and a medium. The system comprises: a data acquisition unit used for acquiring a gastroscope image of a patient, wherein image features of the gastroscope image are represented as fine morphological changes, including clear and visible submucosal blood vessels, easy bleeding and erosion; a model processing unit which is used for carrying out feature extraction on the gastroscope image through a preset recognition network to obtain pathological information, wherein the pathological information comprises a pathological region and a pathological form of a mucous membrane; and a diagnosis output unit which is used for judging whether the patient suffers from gastric cancer or atrophic gastritis according to the pathological information. According to the system, classification, data learning and prediction are performed on the gastroscope images of the atrophic gastritis and the gastric cancer based on the convolutional neural network, accurate diagnosis can be performed on the atrophic gastritis and the gastric cancer, the working efficiency of a gastroenterologist is improved, the diagnosis procedure is simplified, the misdiagnosis and missed diagnosis rate is reduced, and the cost of a patient is reduced.

Description

technical field [0001] The invention relates to the field of medical image processing, in particular to an image recognition system, equipment and medium for distinguishing atrophic gastritis and gastric cancer. Background technique [0002] Gastric cancer is a common malignant tumor of the digestive system. In 2018, there were 1.03 million new gastric cancer cases and 780,000 deaths in the world, the number of cases ranked fifth in the world, and the case fatality rate ranked third. About one-half of the patients are distributed in East Asia, and it is also the second most malignant cancer in my country. Tumor is the largest malignant tumor of the digestive tract (31 / 100,000), and the five-year survival rate was only 27.4%. It is a major and difficult disease that has plagued the masses for a long time. Chronic atrophic gastritis is an independent risk factor and basic condition for gastric cancer. It is an important node and key stage of gastric cancer prevention-intervent...

AI Technical Summary

Problems solved by technology

The gastroscopy of patients with atrophic gastritis is difficult to distinguish from the background mucosa of some early gastric cancer; another study shows that the missed diagnosis rate of gastroscopy doctors with more than 10 years of operating experience is still as high as 10%-20%

In a single electronic gastroscopic examination, the number of pictures taken by the patient usually exceeds 40, and there are many folds in the gastric cavity, and mucus lakes are distributed to block them. All positions in the cavity or coexistence of inflammation and cancer, resulting in missed diagnosis and misdiagnosis by doctors

Therefore, when there are a large number of cases, manual image reading is time-consuming and labor-intensive, and quality control is difficult, which is extremely unfavorable to doctors and patients

[0004] The gold standard for judging atrophic gastritis is pathological examination, which requires at least 5 biopsies. Multiple biopsies increase the risk of mucosal trauma and bleeding. If the patient is taking anti-platelet aggregation drugs such as aspirin or clopidogrel, the risk of bleeding is greater. and increased the economic burden

Images