Title: Analysis of supervised and semi-supervised GrowCut applied to segmentation of masses in mammography images

Abstract: Breast cancer is already one of the most common form of cancer worldwide. Mammography image analysis is still the most effective diagnostic method to promote the early detection of breast cancer. Accurately segmenting tumors in digital mammography images is important to improve diagnosis capabilities of health specialists and avoid misdiagnosis. In this work, we evaluate the feasibility of applying GrowCut to segment regions of tumor and we propose two GrowCut semi-supervised versions. All the analysis was performed by evaluating the application of segmentation techniques to a set of images obtained from the Mini-MIAS mammography image database. GrowCut segmentation was compared to Region Growing, Active Contours, Random Walks and Graph Cut techniques. Experiments showed that GrowCut, when compared to the other techniques, was able to acquire better results for the metrics analyzed. Moreover, the proposed semi-supervised versions of GrowCut was proved to have a clinically satisfactory quality of segmentation.