Automated segmentation and quantitative characterization of radiodense tissue in digitized mammograms.

Mammography has emerged as a reliable non-invasive technique for the early detection of breast cancer — the second leading cause of cancer-related mortality among American women. The radiographic appearance of the female breast consists of radiolucent (dark) regions due to fat and radiodense (light) regions due to connective and epithelial tissue. The amount of radiodense tissue can be used as a marker for predicting breast cancer risk. This paper presents the development of an algorithm for estimating the percentage of radiodense tissue in a digitized mammogram. The technique involves determining a dynamic threshold for segmenting radiodense indications in mammograms. Both the mammographic image and the threshold are modeled as Gaussian random variables. This work is intended to support a concurrent study at the Fox Chase Cancer Center (FCCC) exploring the association between dietary patterns and breast cancer risk. Mammograms have been obtained from an existing cohort of women enrolled in the Family Risk Analysis Program at FCCC. [ABSTRACT FROM AUTHOR]