Potential of ultrasound tomography to measure breast density

Ultrasound tomography (UST), a three-dimensional (3D) whole-breast imaging technology, shows potential to become a radiation-free method of accurately measuring breast density in women of both premammographic and screening age. It could eventually be used to stratify women to different breast cancer screening methods based on their breast density.

Breast density is commonly measured by two-dimensional (2D) mammography quantification. Differences in x-ray attenuation characteristics relate to variations in breast tissue composition, but 2D mammography-derived percent density is subject to error. Magnetic resonance imaging (MRI) is more accurate by providing a 3D volumetric evaluation. MRI density evaluation conducted on T1-weighted sequences using semiautomated segmentation of fibroglandular tissue has demonstrated good correlation with mammography percent density (MPD), as has the Dixon MRI technique, which provides a pure percentage water content of the breast.

Unlike mammography and breast MRI, UST measures the biomechanical tissue properties of mass density and bulk elastic modules. The primary method to assess breast density with UST is by measuring the speed of sound through breast tissue. It improves on mammography by measuring density at each voxel and holds promise as an inexpensive, nonionizing radiation method to evaluate density.

British researchers from the Royal Marsden NHS Foundation Trust and the Institute of Cancer Research in London conducted a study to evaluate the technique of UST and compare whole-breast volume averaged speed of sound (VASS) with percentage water density from noncontrast MR imaging. They evaluated images acquired from 43 healthy volunteers, ranging in age from 30 to 64 years, and reported their findings in Investigative Radiology.>

The women enrolled in the prospective study had a bilateral breast UST (SoftVue, Delphinus Medical Technologies) and a bilateral noncontrast MRI scan. The UST examination was performed with a woman lying prone on the imaging table with a breast suspended in a warm-water bath containing a circumferential transducer array within a uniform ring configuration. Data acquired from many source positions is used to create image reconstructions.<

Two radiologists blinded to the findings of the other were assigned to interpret data from either the UST or the MRI examinations. The researchers created a breast volume of interest (VOI) at a location that was clearly identifiable on both UST and MRI. They manually selected the posterior limit of the VOI as the first coronal frame in which breast tissue was clearly distinct from the chest wall. The anterior limit of the VOI was the last frame before the nipple which did not show a strong reflection signal from the skin. The whole-breast volume averaged speed of sound was calculated by averaging the speed of sound voxel values over the VOI.

MRI image data was acquired by using an axial bilateral proton tensity-weighted 2-point Dixon sequence at high resolution. The VOI used the same positioning rules. The water fraction was calculated for every voxel within the VOI and the percentage of water content was measured.

Lead author Elizabeth A. M. O’Flynn, MD, and colleagues reported a very high similarity between measurements obtained from both the left and right breasts, and that the VASS from UST was highly reproducible. For MRI, there was a small but significant difference in MR percent water content between the left (35.1%) and right (36.3%) breasts.

The researchers identified a strong association of VASS with MR percent water content. They stated that this finding indicates that VASS could be a potentially alternative surrogate 3D measure of breast density. Since sound speed is more directly linked to the physical density of breast tissue, it has the potential to be more accurate and more relevant than MPD as a measure of breast density.

REFERENCE

  1. O’Flynn EAM, Fromageau J, Ledger AE, et al. Ultrasound Tomography Evaluation of Breast Density. Invest Radiol. 2017 52; 6: 343-348.
© Anderson Publishing, Ltd. 2024 All rights reserved. Reproduction in whole or part without express written permission Is strictly prohibited.