Segmentation of the Breast Region in Mammograms Using Snakes
CRV '04 Proceedings of the 1st Canadian Conference on Computer and Robot Vision
Breast segmentation with pectoral muscle suppression on digital mammograms
IbPRIA'05 Proceedings of the Second Iberian conference on Pattern Recognition and Image Analysis - Volume Part II
Hi-index | 0.00 |
The segmentation of a digital mammogram into breast region and film background is a necessary prerequisite in the computer-assisted di- agnosis of mammograms. The segmentation method should be ro- bust enough to handle a wide variety of digital mammograms obtained from different image acquisition systems. In addition, the large size of digitized mammograms requires an efficient image processing system. In this paper, a sofware package called United Snakes for interactive medical image segmentation is evaluated for segmenting the breast re- gion from mammograms. The united snakes technique combines the livewire method with a group of most important snake variants united within a general framework. The advantages of the system are general- ity, accuracy, ease of use and robustness. Practical tests showed that the method gives an accurate segmentation result with only few user interactions.