Jun Ge PhD, Presenter: Employee, Hologic, Inc

Ashwini Kshirsagar, Abstract Co-Author: Employee, Hologic, Inc

Julian Marshall, Abstract Co-Author: Employee, Hologic, Inc

Haili Chui PhD, Abstract Co-Author: Employee, Hologic, Inc

Kevin Kreeger, Abstract Co-Author: Employee, Hologic, Inc

Zhenxue Jing, Abstract Co-Author: Employee, Hologic, Inc

Studies of correlation between breast arterial calcium (BAC) and risk of coronary artery disease, diabetes mellitus and other diseases in women have yielded mixed conclusions. Most clinical studies used subjective methods and categorical scales to measure the presence of breast arterial calcium. Such measurement methods may suffer from inter- and intra-reader variability and produce inaccurate results. The purpose of this study is to evaluate the accuracy of an automated method for objectively measuring BAC using digital mammograms.

Four standard mammographic views of 311 consecutive BIRADS I cases were collected from three clinics. Cases with surgical clips, implants or pace-makers were excluded from this study. The “truth” for arterial calcium was determined using an interactive segmentation tool by manually delineating segments of calcified arteries on digital mammographic images. These images were then analyzed using a Calcium Scoring prototype software to automatically measure the amount of arterial calcium.

20 out of the 311 cases (80 images) exhibited calcified arteries. If we presume that an area threshold of 10 mm2 is used to separate cases into presence of clinically significant BAC or not then 60 of the 80 images would be considered normal from the truth marking. Of these 60, the Calcium Scoring prototype software identified less than 10 mm2 BAC in 59 of them (specificity = 98%). Of the 20 images with more than 10 mm2 BAC in truth marking the Calcium Scoring prototype software identified more than 10 mm2 BAC in 15 of them (sensitivity=75%).

An automated BAC measurement tool provided an accurate and objective method to measure total amount of breast arterial calcium in digital mammograms.

An automated BAC quantification tool has been developed for clinical research in evaluating the correlation between extent of breast arterial calcifications in digital mammograms and risk of diseas

Ge, J, Kshirsagar, A, Marshall, J, Chui, H, Kreeger, K, Jing, Z, Quantification of Breast Arterial Calcium Using Digital Mammograms.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9012604.html