Related Experiment Videos
A segmentation technique to detect masses in dense breast digitized mammograms
Vivian Toledo Santos1, Homero Schiabel, Cláudio E Góes
1Departamento de Engenharia Eléctrica, Escola de Engenharia de São Carlos-SP, Brazil. vivian@jaunet.com.br
Journal of Digital Imaging
|July 10, 2002
Summary
This study presents a watershed transformation technique for detecting masses in mammograms. The method achieved 85.4% true positive detection, offering a promising tool for breast cancer screening.
Area of Science:
- Medical Imaging
- Computer-Aided Diagnosis
- Biomedical Engineering
Background:
- Mammography is crucial for early breast cancer detection.
- Dense breast tissue can obscure masses, challenging interpretation.
- Automated mass detection aims to improve screening accuracy.
Purpose of the Study:
- To develop and evaluate a novel segmentation technique for mass detection in digitized mammograms.
- To assess the efficacy of watershed transformation for identifying suspicious regions in dense breast tissue.
Main Methods:
- A four-step technique involving preprocessing, histogram equalization, watershed transformation, and topography.
- Processing 109 regions of interest from mammograms.
- Utilizing watershed transformation for gradient calculation and segmentation.
Main Results:
- The technique achieved 85.4% true positive detection and 20% false positive detection on equalized images.
- Preliminary results indicate good performance in mass detection.
- Identified need for further noise reduction techniques for certain images.
Conclusions:
- The watershed transformation technique shows potential as a valuable tool in mammography for mass detection.
- Further development is needed to address image noise and improve overall accuracy.
- The method contributes to advancing computer-aided diagnosis in breast imaging.