This technology is a fully automated computer algorithm used to estimate volumetric breast density from 3D digital breast tomosynthesis (DBT) images.
Increased breast density is a risk factor for developing breast cancer; therefore, accurately measuring breast density is important for informing disease risk assessments and guiding patient screening recommendations. Current methods for classifying breast density rely on subjective, qualitative visual assessments with minimal reproducibility across multiple observers. While magnetic resonance imaging (MRI) and full-field digital mammography (FFDM) can be used to estimate volumetric breast density, these approaches are limited by their availability and the 2D nature of FFDM, respectively. Therefore, objective, quantitative, and accurate methods for estimating breast density that utilize accessible, 3D digital breast tomosynthesis (DBT) images are needed.
This technology is a fully automated method for quantifying volumetric breast density from digital breast tomosynthesis (DBT) images. This technology utilizes a computer algorithm to identify dense areas of the breast across multiple different DBT images. These images can then be modeled into a predictive 3D model structure in order to quantify volumetric breast density. This method has shown significant correlation with estimates measured using established methods to quantify volumetric breast density, including FFDM, MRI, and manual evaluation.
This technology has been validated with human breast DBT images.
IR CU26093
Licensing Contact: Joan Martinez
