Quantitative Analysis of Breast Images

M. Kallenberg

• Promotor: N. Karssemeijer
• Graduation year: 2012
• Radboud University, Nijmegen

Abstract

Breast density both seriously impairs the sensitivity of mammography and increases the risk for developing breast cancer. As such it is essential to accurately measure breast density. Measuring breast density is, however, not a trivial task, as a quantitative assessment is time consuming and observer dependent. In this thesis we have developed and improved techniques to automatically and accurately measure breast density in breast images of different modalities. We demonstrated that the developed methods showed good correspondence to human observers. As such, we have shown that the automatic measurement of breast density is feasible, thereby providing opportunities to optimize breast cancer screening.