Related Experiment Video
Updated: May 30, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
Effectiveness of computer-aided detection in community mammography practice
Joshua J Fenton1, Linn Abraham, Stephen H Taplin
1Department of Family and Community Medicine and Center for Healthcare Policy and Research, University of California, Davis, Sacramento, CA 95817, USA. joshua.fenton@ucdmc.ucdavis.edu
Journal of the National Cancer Institute
|July 29, 2011
Summary
Computer-aided detection (CAD) in mammography reduced specificity and positive predictive value without improving invasive breast cancer detection or prognostic characteristics. Further research is needed to optimize CAD performance.
Area of Science:
- Radiology and Medical Imaging
- Oncology
- Public Health
Background:
- Computer-aided detection (CAD) is widely used in screening mammography in the US.
- Its impact on breast cancer detection by community radiologists remains uncertain.
- Millions of women undergo screening mammography annually with CAD.
Purpose of the Study:
- To investigate the association between CAD use and mammography performance metrics.
- To assess CAD's impact on breast cancer detection rates and prognostic characteristics.
- To evaluate CAD's effectiveness in film-screen screening mammography.
Main Methods:
- Analysis of over 1.6 million film-screen mammograms from 684,956 women across US facilities (1998-2006).
- Used random-effects logistic regression to adjust for patient and examination factors.
- Examined associations between CAD use and specificity, sensitivity, and positive predictive value.
Main Results:
- CAD use was associated with significantly lower specificity (OR=0.87) and positive predictive value (OR=0.89).
- Overall sensitivity showed a non-significant increase with CAD (OR=1.06), with higher sensitivity for ductal carcinoma in situ but not invasive cancer.
- CAD did not improve breast cancer detection rates or the prognostic characteristics (stage, size, node involvement) of invasive cancers.
Conclusions:
- Computer-aided detection use in film-screen mammography is linked to decreased specificity.
- CAD does not enhance the detection rate or improve prognostic features of invasive breast cancer.
- The study highlights potential drawbacks of CAD in current screening mammography practices.
