Having a dense breast tissue (heterogeneously or extremely dense) has been linked to an increased risk of dveloping breast cancer. Breast density is usually routinely measured in mammography by comparing the ratio of stromal tissue to fatty tissue. Because increased breast density is associated with a higher risk for breast cancer, therapies that reduce breast density have been proposed as cancer preemptive treatments. For studies assessing the effect of breast density reduction therapies, it is desirable to follow changes in breast density longitudinally. Unfortunately, the radiation exposure in mammography makes the technique impractical for serial studies of breast density. Also, since mammograms yeild two-dimensional information, the breast densities may change based on the projection, level and angle of compression, and scanner calibration. All these reasons makes digital mammograhy not very suitable and reliable for breast density measurement.

Magnetic resonance imaging (MRI) on the other hand is a noninvasive three dimensional imaging technique that uses nonionizing radiation. MRI yields distinct fat and water signals. Also, it does not expose the subject to ionizing radiation and can capture time-varying features in real time effectively. Thus for longitudinal studies of patients at risk, breast density can be better assessed using MRI. In recent years, several MRI techniques have been developed for the measurement of fat-water content. In this study, we use one of these techniques: the radial gradient-echo and spin-echo (RADGRASE). This technique yields fat, water, and T2-corrected fat-fraction images for the entire breast from data acquired within few minutes.

In order to analyze the breast density in the breast, the first step is to define a region of interest (ROI) within the MR images by segmenting the breast from the rest of the image. The manual segmentation of a stack of breast image slices is tedious, time consuming and impractical. Due to the noise and unrelated high-contrast edges and regions within the MR images, automatic segmentation of the correct breast surfaces is a challenging task.

In this research, we developed an automated system capable of localizing and extracting the region of interest (breast tissues) within the fat and water MR images --- i.e., we developed a 3-D image segmentation algorithm capable of segmenting the breast tissues within these MR images.

The video on the top left shows a typical sequence of fat MR images captured. The image sequence on the bottom right shows the result obtained after each step of the segmentation algorithm on a single frame from this video.

This work was a collaborative effort with Prof. María I. Altbach in the Dept. of Medical Imaging, College of Medicine, University of Arizona.

Publications:

1. José A. Rosado-Toro, Tomoe Barr, Jean-Phillipe Galons, Marilyn T. Marron, Alison Stopeck, Cynthia Thomson, Patricia Thompson, Danielle Carroll, Eszter Wolf, María Altbach, and Jeffrey J. Rodriguez "Automated Breast Segmentation of Fat and Water MR Images Using Dynamic Programming," Academic Radiology, vol. 22, no. 12, Feb 2015, pp. 139-148. [ PDF ]