Mammographic density is one of the strongest predictors of breast cancer risk, second only to carrying a mutation in the BRCA1 or BRCA2 genes.
Mammographic density is the white area on a mammogram representing the radiographic appearance of epithelial and stromal tissue (as opposed to fatty tissue which appears dark on a mammogram). It cannot be determined by feel or touch. Women with extensive mammographic density are 4-6 times more likely to develop breast cancer than women of similar age with little or no mammographic density.
There is enormous potential for screening programs, like BreastScreen, to use measurements of mammographic density to identify and target women at higher risk of disease. Many screening programs have transitioned from film to digital mammography and whilst the overall diagnostic accuracy of film and digital mammography as a means of screening for breast cancer is similar, differences in the relative amounts of mammographic density shown in a traditional film compared to a digital image has not been assessed. The measurement of mammographic density from digital mammograms may also be affected by different processing algorithms applied to the digital images by different manufacturers of mammography machines. Without adjustment, these differences could introduce bias to dozens of epidemiological studies that rely on a mixture of measurements from different types of mammograms. Newly developed automated methods of mammographic density designed for clinical use will also need to be validated on different types of mammograms.
The specific aims of this project are to:
- Derive calibration algorithms for comparison of mammographic density measurements in film and digital mammograms so that direct comparisons can be made
- Investigate the effects of processing applied by different digital mammography manufacturers.
This project is a collaborative effort between GOHaD and the University of Melbourne:
There are currently several research groups interested in incorporating mammographic density measurements into breast cancer risk models however the heterogeneity of mammographic density measurements due to mammogram type, processing, and measurement technique, makes accurately predicting an individual women’s risk impossible. This research will help describe potential sources of bias and provide a resource for testing and validation.
For more information, contact Jennifer Stone.