Mammographie mass classification based on active contour

Computer-aided diagnosis is being developed to assist radiologists in the interpretation of mammograms. This could represent to further amelioration by increasing diagnostic accuracy in the screening programs. This research have developed a computerized scheme for segmentation and classification of mass in digital mammograms, that were digitized in the acquisition phase. After the digitization process, the breast border was initially determined. An active contour algorithm was employed for mass boundary segmentation. Finally, percentage of mammographic density was calculated for mass classification according to the fourth edition of BI-RADS lexicon. The sensitivity achievement was 88% for mass classification.