Method for measuring content of glands in breast tissues based on magnetic resonance image histogram threshold segmentation technology

Abstract

The invention discloses a method for measuring the content of glands in breast tissues based on a magnetic resonance image histogram threshold segmentation technology. The method comprises the steps of firstly performing magnetic resonance scanning on a testee to obtain a magnetic resonance image; secondly loading a T1 weighted image sequence in the magnetic resonance image into an image postprocessing program, and displaying signal intensity distribution of voxels in T1 weighted images by using a histogram; thirdly filtering surrounding air noises irrelevant to detection, and then removing other regions irrelevant to the breast tissues, thereby obtaining complete double-side breast tissues; fourthly according to signal characteristics of gland tissues and adipose tissues, separating the gland tissues and the adipose tissues in the breasts; and finally calculating the total volume of the breast tissues and the volume of the gland tissues, namely, obtaining a volume ratio of the gland tissues in the whole breast tissues. According to the method, the percentage of the gland tissue volume in the whole breast tissue volume can be intuitively obtained; and a result is more objective andaccurate.

Description

technical field [0001] The invention relates to the detection of gland content in breast tissue, in particular to a method for measuring the gland content in breast tissue based on a magnetic resonance image histogram threshold segmentation technique. Background technique [0002] Breast examination is often involved in routine gynecological examinations. Breast imaging examinations mainly include mammography, ultrasound, and magnetic resonance (magnetic resonance, MR) examinations. These examinations lack standardized and unified standards for breast classification. The Breast Imaging Reporting and Data System (BI-RADS) developed by the American College of Radiology divides the breast into four types: adipose type (almost all of the breast is adipose tissue, glandular tissue <25%), a small amount of glandular type (breast Scattered glandular tissue, accounting for 25-50%), a large amount of glandular type (the breast is unevenly dense, and the glandular tissue accounts ...

AI Technical Summary

Problems solved by technology

[0003] At present, there are few ultrasound reports on breast typing, and ultrasound typing generally selects the upper outer quadrant of the breast to observe the thickness and glandular structure of the breast parenchyma, and generally selects the thickest part of the upper outer quadrant to measure the thickness of the gland. The range is relatively single, the three-dimensional effect is poor, and it is greatly affected by the operator's technique

MRI examination classification is similar to mammography classification. The ratio of area and volume is selected as the basis of classification. Although the three-dimensional sense is enhanced compared with ultrasound, the ratio of breast glands and fat is still judged by the naked eye. The measurement standard Not standardized and comprehensive enough

Since there is still a lack of standardized and unified standards for the methods and terminology of breast typing in imaging, the breast types obtained in the study are not very referable

Images