Related Experiment Video
Updated: Apr 26, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
42.6K
Reduction of false-positive recalls using a computerized mammographic image feature analysis scheme
Maxine Tan1, Jiantao Pu, Bin Zheng
1School of Electrical and Computer Engineering, University of Oklahoma, Norman, OK 73019.
Physics in Medicine and Biology
|July 18, 2014
Summary
A new computer-aided diagnosis (CAD) scheme analyzing mammogram texture and density features significantly improved cancer detection accuracy. This approach may help reduce false-positive recalls in mammography screening.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Radiology
Background:
- High false-positive recall rates in mammography screening reduce efficacy and increase healthcare costs.
- Accurate differentiation between benign and malignant cases among recalled women is crucial.
Purpose of the Study:
- To develop and assess a computer-aided diagnosis (CAD) scheme for reducing false-positive recalls in mammography.
- To analyze global mammographic texture and density features from four-view images.
Main Methods:
- Utilized a database of 1052 full-field digital mammography (FFDM) cases (669 cancer, 383 benign).
- Computed global texture and density features from craniocaudal (CC) and mediolateral oblique (MLO) views.
- Employed artificial neural network (ANN) classifiers and an adaptive scoring fusion method.
Main Results:
- The four-view CAD scheme achieved an area under the receiver operating characteristic curve (AUC) of 0.793 ± 0.026.
- This performance was significantly higher than using CC or MLO views alone (p < 0.05).
Conclusions:
- Quantitative assessment of mammographic texture and density features aids in classifying recalled cases.
- The developed CAD scheme shows potential for reducing false-positive recalls in screening mammography.
