Lesion area quantification method based on mammary gland molybdenum target image

Abstract

The invention relates to a lesion area quantification method based on a mammary gland molybdenum target image, and the method comprises the steps: carrying out the detection network training of a mammary gland lesion area through an efficientdet-d8 network and a YOLOV5 network, and obtaining a detection model; performing breast lesion area segmentation training through a Unet network structure, performing two Unet network structure training according to the size of the breast molybdenum target image, performing segmentation network training on a large lesion area, and training a segmentation model on a small lesion area; after the detection model and the segmentation model are obtained, carrying out verification respectively; calculating the position of a lesion area through a detection model and a segmentation model according to an acquired mammary gland molybdenum target image, calculating the size and density distribution of the lesion area through a segmentation result, obtaining a benign and malignant confidence score through a detection result, and providing diagnosis assistance information through positioning and quantification of the lesion area. The problem that human experience is insufficient when breast diseases are read can be solved, and doctors are helped to locate and analyze the diseases.

Description

technical field [0001] The patent of the present invention relates to the field of medical aided diagnosis, in particular to a method for quantifying lesion areas of breast diseases. Background technique [0002] With the development of society and economy, people's lifestyles have undergone great changes, such as early sexual maturity, delayed menopause, infertility, late childbirth, non-breastfeeding, anxiety, bad mood, obesity, intake of high-calorie and high-fat diet Income, unhealthy lifestyle and other living environment factors have increased the incidence of breast diseases year by year, among which breast cancer is the most prominent. Breast cancer is one of the most frequent and common malignant tumor diseases in women, which poses a serious threat to women's physical and mental health and quality of life. . Breast cancer ranks first in global cancer incidence and fifth in mortality. In recent years, the incidence and mortality of breast cancer in China have been...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method for quantifying the lesion area based on mammography images, which is used to solve the problem that the prior art requires a higher level of diagnosis for doctors

Images