Classification of mass and non-mass mammography based on Tsallis entropy

Breast cancer is a neoplasia that affects a high number of women in the world every year and it is the second on ranking of the diseases that more affects women. Mammography is the most important exam to aid in early detection and diagnosis of breast cancer. Computational methods have been developed to assist the radiologist in diagnosis and to improve the perception of the results. Feature extraction of mammography images has been allowed good results in classification of these mass. This paper presents a new approach of texture descriptors using multilevel decomposition and extracting Tsallis entropy from this new images and the best result for a vector of 24 features was 84.21% of accuracy.