Related Experiment Videos
Computer-aided detection of clustered microcalcifications in digitized mammograms
B Zheng1, Y H Chang, M Staiger
1Department of Radiology, University of Pittsburgh, PA 15261-0001, USA.
Academic Radiology
|August 1, 1995
Summary
A new computer-aided detection (CAD) scheme for clustered microcalcifications in mammograms achieved 100% sensitivity. This advanced CAD system shows promise for improving early breast cancer detection accuracy.
Area of Science:
- Medical Imaging
- Radiology
- Computer-Aided Diagnosis
Background:
- Early detection of breast cancer is crucial for effective treatment.
- Microcalcifications are common indicators of early breast cancer on mammograms.
- Computer-aided detection (CAD) systems aim to improve the accuracy of mammographic interpretation.
Purpose of the Study:
- To investigate the performance of a novel computer-aided detection (CAD) scheme.
- To evaluate the CAD scheme's effectiveness in detecting clustered microcalcifications in digitized mammograms.
Main Methods:
- Development of a multistage CAD scheme incorporating a Gaussian band-pass filter and nonlinear threshold for enhanced sensitivity.
- Implementation of a local minimum searching routine and multilayer topographic feature analysis to minimize false positives.
- Testing the CAD scheme on 110 digitized mammograms, including 55 with verified microcalcification clusters.
Main Results:
- The developed CAD scheme demonstrated 100% sensitivity in detecting clustered microcalcifications.
- The system achieved a low average false-positive detection rate of 0.18 per image.
- The CAD scheme's performance is comparable to existing published systems.
Conclusions:
- The CAD scheme exhibits high sensitivity and a low false-positive rate for detecting clustered microcalcifications.
- This CAD system offers unique advantages that can potentially enhance future detection sensitivity and specificity.
- The findings suggest the CAD scheme is a valuable tool for improving mammographic analysis.