Paper
10 March 2017 Exploring a new bilateral focal density asymmetry based image marker to predict breast cancer risk
Author Affiliations +
Abstract
Although breast density has been widely considered an important breast cancer risk factor, it is not very effective to predict risk of developing breast cancer in a short-term or harboring cancer in mammograms. Based on our recent studies to build short-term breast cancer risk stratification models based on bilateral mammographic density asymmetry, we in this study explored a new quantitative image marker based on bilateral focal density asymmetry to predict the risk of harboring cancers in mammograms. For this purpose, we assembled a testing dataset involving 100 positive and 100 negative cases. In each of positive case, no any solid masses are visible on mammograms. We developed a computer-aided detection (CAD) scheme to automatically detect focal dense regions depicting on two bilateral mammograms of left and right breasts. CAD selects one focal dense region with the maximum size on each image and computes its asymmetrical ratio. We used this focal density asymmetry as a new imaging marker to divide testing cases into two groups of higher and lower focal density asymmetry. The first group included 70 cases in which 62.9% are positive, while the second group included 130 cases in which 43.1% are positive. The odds ratio is 2.24. As a result, this preliminary study supported the feasibility of applying a new focal density asymmetry based imaging marker to predict the risk of having mammography-occult cancers. The goal is to assist radiologists more effectively and accurately detect early subtle cancers using mammography and/or other adjunctive imaging modalities in the future.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Faranak Aghaei, Seyedehnafiseh Mirniaharikandehei, Alan B. Hollingsworth, Yunzhi Wang, Yuchen Qiu, Hong Liu, and Bin Zheng "Exploring a new bilateral focal density asymmetry based image marker to predict breast cancer risk", Proc. SPIE 10136, Medical Imaging 2017: Image Perception, Observer Performance, and Technology Assessment, 101361P (10 March 2017); https://doi.org/10.1117/12.2254073
Lens.org Logo
CITATIONS
Cited by 1 scholarly publication.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Cancer

Mammography

Computer aided diagnosis and therapy

Breast cancer

Image segmentation

Breast

Computer aided design

Back to Top