Endometrial cancer cell detection method, system and equipment and storage medium

Abstract

The invention discloses an endometrial cancer cell detection method, system and equipment and a storage medium. The method comprises the following steps: preparing an endometrial cell pathological section; preliminarily screening images containing positive cells by using CNN; performing intelligent detection on cancer cells by using the improved Faster R-CNN target detection model; and performing statistics on each index, sending the WSI generated blocks into network prediction, and then merging identification results. In the target acquisition part, a data set is normalized and expanded by using a self-adaptive random clipping method on the aspect of data; in the model, an improved CNN model and a two-stage Faster RCNN model are taken as basic models, diagnosis experience of doctors is introduced, and diagnosis indexes are extracted to jointly act to train a more accurate model. The accuracy rate of the endometrial cancer cell detection method is higher than that of a classification method, meanwhile, the confidence coefficient and the position information of the WSI can be given, and in addition, the invention further provides a preparation method of a more effective endometrial cell pathological section image data set. The method is applied to the field of medical image target detection.

Description

technical field [0001] The invention belongs to the technical field of AI medical imaging, and relates to an endometrial cancer cell detection method, system, equipment and storage medium. Background technique [0002] Endometrial cancer is a group of epithelial malignant tumors that occur in the endometrium. It is one of the three major malignant tumors of the female reproductive system. There are more than 378,000 patients in the world every year, and its incidence is increasing worldwide. and tend to be younger. In addition, studies have shown that the 5-year survival rate for early-stage endometrial cancer with appropriate treatment is over 85%. Early screening and early diagnosis of precancerous lesions of endometrial cancer can greatly reduce its incidence and mortality. [0003] Computer-aided diagnosis systems can help doctors effectively analyze and evaluate a large number of medical images. It has been applied to the radiodiagnosis of some common cancers such as ...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problems of slow time-consuming, low precision and strong subjectivity in the pathological image diagnosis process of endometrial cancer cells, and provide a method, system, equipment and storage medium for detecting endometrial cancer cells

Images