Preprocessing method and preprocessing system used for extracting breast regions in mammographic images

Abstract

The invention discloses a preprocessing method and a preprocessing system used for extracting breast regions in mammographic images. The method includes steps: subjecting the mammographic images to automatic threshold segmentation to obtain binary images; subjecting communicated regions in the binary images to filling processing to remove labels and holes in the mammographic images; dividing the binary images into a plurality of subregions, and binarizing each subregion according to local threshold value to position breast contour regions; subjecting the binary images and the breast contour regions to set 'and' operation, searching number of pixel points in each subregion containing difference in the binary images along a reference line from top to bottom, and positioning the subregion with maximum number of pixel points in regions containing difference to be nipple regions. By overall automatic threshold segmentation and local threshold analysis, the breast regions and the nipple regions in the mammographic images can be accurately positioned, the preprocessing process is low in computation complexity, and execution efficiency is high.

Description

technical field [0001] The invention relates to a medical image processing method, in particular to a preprocessing method and system for extracting mammary gland regions from mammogram images. Background technique [0002] The clarity and contrast of the original mammography images (mammography) are often insufficient. In order to help clinicians better diagnose the disease, the original images need to be processed and enhanced. However, mammography images have background, label and other interference information, and different breast regions have different image properties, so different enhancement algorithms need to be used for processing. Therefore, it is necessary to preprocess and segment the input raw breast images. [0003] Most of the current mammography image segmentation algorithms are simple single-threshold segmentation, which results in poor accuracy of breast regions and rough edges. [0004] The Chinese patent application with application number 20091010065...

AI Technical Summary

Problems solved by technology

[0006] The present invention proposes a preprocessing method and system for extracting mammary gland regions from mammogram images, by removing the background and labels of the mammogram images, using global automatic threshold segmentation and local threshold analysis to accurately locate mammogram images The mammary gland area and the nipple area are obtained to solve the current technical problem that the nipple area cannot be accurately located

Images