Method, device and computer storage medium for detecting lesion area of ​​breast image

Abstract

The invention discloses a method for detecting lesion areas of mammary gland images. The method includes: receiving a mammary gland image to be detected, and preprocessing the mammary gland image to be detected; performing a clustering and segmentation on the preprocessed mammary gland image based on a Nystrom spectral clustering algorithm to obtain a suspicious mammary gland lesion area; The K-means clustering algorithm performs secondary clustering and segmentation on the suspicious mammary gland lesion region to obtain a corresponding region of interest; extracts feature information of the region of interest, and detects whether the region of interest is Breast lesion area. The invention also discloses a detection device and a computer-readable storage medium for the mammary gland image lesion area. The invention can improve the accuracy of segmenting lesion areas in mammary gland images, thereby improving the accuracy of detection results of mammary gland lesion areas.

Description

technical field [0001] The present invention relates to the technical field of medical image processing, in particular to a method, device and computer storage medium for detecting lesion areas of mammary gland images. Background technique [0002] Breast cancer is a common malignant tumor, early diagnosis and treatment is the key to reduce breast cancer mortality. The computer-aided detection system can help doctors make final diagnostic decisions by detecting suspicious lesion areas, thereby improving the survival rate and quality of life of breast cancer patients. Since lumps and calcification clusters are the most common imaging signs of breast cancer, the automatic detection of lumps and calcifications has also become the two main aspects of the computer-aided diagnosis system. Among them, tumors have always been a difficulty in computer-aided detection due to factors such as blurred edges, various shapes, and low contrast with surrounding tissues. Especially for dens...

AI Technical Summary

Problems solved by technology

Among them, the first) method is only based on the feature of breast lesion area density, which is not good for the detection of lesion areas in dense breast images, and the second) method only relies on the K-means clustering algorithm to extract interest Regions, the segmentation effect is better for circular or quasi-circular lesions with clearer edges and more regular shapes, but for lesions with irregular shapes and hidden in dense tissue regions, more accurate tumor edges cannot be obtained.

Therefore, the existing technology cannot accurately segment the lesion area in the dense breast image, resulting in poor detection results

Images