Breast imaging plays an outstanding role for the diagnosis of asymptomatic women through micro-calcification (μCa) detection. This paper is devoted to mammographic images analysis related to the study of methods for automatic μCa characterization in breas

C. QUINTANA; M. VALENTE; G. TIRAO

COMPUTERS IN BIOLOGY AND MEDICINE

PERGAMON-ELSEVIER SCIENCE LTD

Año: 2012 p. 1 - 37

0010-4825

Breast imaging plays an outstanding role for the diagnosis of asymptomatic women through micro-calcification (μCa) detection. This paper is devoted to mammographic images analysis related to the study of methods for automatic μCa characterization in breast imaging, based on image segmentation and object recognition. These methods were applied to a set of simulated images considering a wide range of clinical situations. The registration algorithm has shown good performance to μCa localization, even by means of simple and absolutely automatic methods. Regarding μCa characterization, thresholding technique produced results with a high concordance with μCa, regardless of their shape or size.