Tsallis Entropy based Microcalcification Segmentation

This paper proposes a new segmentation approach by considering the Non extensive property of mammograms. The novel thresholding technique is performed by Tsallis entropy characterized by one more parameter q, which depends on the nonextensiveness of mammogram. Mammograms are typical examples of image with fractal-type structures (nonextensiveness).The proposed approach has been tested on 388 mammograms, and the results have demonstrated that the proposed Tsallis fuzzy approach outperforms the 2D nonfuzzy approach and traditional Shannon entropy (SE) partition approach. Some typical results are presented to illustrate the influence of the parameter q in the thresholding.