Automated Method for Analysis of Mammographic Breast Density - A Technique for Breast Cancer Risk Estimation
Abstract
The goal of this proposed project is to develop an automated technique to assist radiologists in estimating mammographic breast density. During this project year, we have completed the analysis of the correlation between the percent mammographic dense area and the percent volumetric fibroglandular tissue measured on MR images. The performance of the automated segmentation program on the Set of mammograms used in this study was verified with an experienced radiologist's manual segmentation. The percent mammographic dense area and percent volumetric fibroglandular tissue is highly correlated with correlation coefficients of 0.91 and 0.89, respectively, for CC and MLO views. The high correlation indicates the validity of using mammographic density as a surrogate for monitoring breast density changes. The computerized image analysis tool can provide a consistent and reproducible estimation of percent dense area on routine clinical mammograms, thereby contributing to the understanding of the relationship of mammographic density to breast cancer risk, detection, and prognosis, and the prevention and treatment of breast cancer. We have also begun the development of an automated density segmentation method for direct digital mammograms (DMs). We performed a study to compare breast density estimated from digitized screen-film mammograms (SEMs) with that estimated from DMs. Our results indicate that breast density on DMs generally appears to be lower than that on SFMs because of the harder beam quality used and image processing applied to the DMs. The lower density may improve the mammographic sensitivity for lesion detection in dense breasts. However, for patients with SFMs and DMs taken over time, comparison of serial mammograms for breast density changes will be problematic. We are designing automated segmentation techniques for DM1. We will modify the program and evaluate its performance on DMs in the coming year.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 2003
- Accession Number
- ADA418148
Entities
People
- Heang P. Chan
Organizations
- University of Michigan