The application's broad, long-term objective is to reduce the disease burden of breast cancer through early detection and accurate diagnosis, which lead to effective treatment. The goal of this project is to develop innovative approaches to computer-aided diagnosis (CADx) of breast calcifications for breast cancer detection and diagnosis. Other researchers in our group are developing CADx for breast masses separately, in conjunction with this project, because diagnostic workup of breast masses requires multi-modality imaging (ultrasound and MRI) in addition to mammography. We will test the hypothesis that CADx of breast calcifications can help improve breast cancer detection and diagnosis.
The Specific Aims of this project are: (1) reduce or eliminate influence on computer classification from inter- and intra-radiologist variability in their interaction with the computer;(2) investigate computer classification of magnification mammogram;(3) investigate and reduce computer variability in classification of a lesion in multiple views;and (4) investigate potential benefit of CADx to enhance the effectiveness of computer-aided detection (CADe) in screening mammography. The research design will be to understand the mechanisms of computer variability in its calculations from the identified sources, to develop new techniques to minimize computer calculation variability, to develop new techniques for new CADx applications, and to perform an observer performance study to show that CADx can potentially enhance the effectiveness of CADe. The methods to be used include computer image analysis, statistical classifier, ROC analysis, statistical comparison, and observer performance study. The rationales for pursuing these goals include an anticipated need to address critical clinical acceptance issues of CADx based on already demonstrated high performance of our CADx technique and a planned pre-clinical and clinical evaluation of CADx, and a new opportunity for potential novel application of CADx to enhance the effectiveness of CADe. The techniques that we will use are either proven in previous investigations, or based on theoretical studies, or based on our observation in working with radiologists developing CADx techniques. The importance and health relatedness of the research described in this application is that it will address significant limitations of current CADx technique that likely will hinder clinical acceptance of CADx and will address an innovative potential for CADx to enhance the effectiveness of CADe in breast cancer detection. If the Aims of the application are achieved, CADx will be advanced technologically and will become more clinically acceptable to radiologists: once clinical benefit of CADx is demonstrated, CADx will become an important new clinical tool for breast cancer detection and diagnosis.

Public Health Relevance

The goal of this project is to develop innovative approaches to computer-aided diagnosis (CADx) of breast calcifications for breast cancer detection and diagnosis. This project will address significant limitations of current CADx technique with the objective of advancing CADx to become a clinical tool for breast cancer detection and diagnosis.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Research Project (R01)
Project #
Application #
Study Section
Biomedical Imaging Technology Study Section (BMIT)
Program Officer
Zhang, Yantian
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Chicago
Schools of Medicine
United States
Zip Code
Soylu, Fatma Nur; Peng, Yahui; Jiang, Yulei et al. (2013) Seminal vesicle invasion in prostate cancer: evaluation by using multiparametric endorectal MR imaging. Radiology 267:797-806
Schmid-Tannwald, Christine; Jiang, Yulei; Dahi, Farid et al. (2013) Diffusion-weighted MR imaging of focal liver lesions in the left and right lobes: is there a difference in ADC values? Acad Radiol 20:440-5
Peng, Yahui; Jiang, Yulei; Yang, Cheng et al. (2013) Quantitative analysis of multiparametric prostate MR images: differentiation between prostate cancer and normal tissue and correlation with Gleason score--a computer-aided diagnosis development study. Radiology 267:787-96
Jiang, Yulei (2013) On the shape of the population ROC curve. Acad Radiol 20:897-907
Liu, Bei; Jiang, Yulei (2013) A multitarget training method for artificial neural network with application to computer-aided diagnosis. Med Phys 40:011908
Jiang, Yulei; Metz, Charles E (2010) BI-RADS data should not be used to estimate ROC curves. Radiology 256:29-31
Zur, Richard M; Pesce, Lorenzo L; Jiang, Yulei (2010) The effect of two priors on Bayesian estimation of "Proper" binormal ROC curves from common and degenerate datasets. Acad Radiol 17:969-79
Zur, Richard M; Jiang, Yulei; Pesce, Lorenzo L et al. (2009) Noise injection for training artificial neural networks: a comparison with weight decay and early stopping. Med Phys 36:4810-8
Krupinski, Elizabeth A; Jiang, Yulei (2008) Anniversary paper: evaluation of medical imaging systems. Med Phys 35:645-59
Rana, Rich S; Jiang, Yulei; Schmidt, Robert A et al. (2007) Independent evaluation of computer classification of malignant and benign calcifications in full-field digital mammograms. Acad Radiol 14:363-70

Showing the most recent 10 out of 16 publications