Résumé:
Breast cancer continues to be one of the main causes of female mortality in the world of
today. For this reason several works is carried out, in order to develop tools of assistance with
the diagnosis of this disease, each one having its advantages and its operational limits. The
treatment carried out within the framework of this memory, concentrates on the extraction of
calcifications by techniques of mathematical morphology, adapting to the structures
mammaires, in order to evaluate texture descriptors. These discriminates, are integrated in a
classifier, carried out by the technique of the decision trees, in order to distinguish
calcifications in benign matter and those, in malignant matter.
Key words: mammographic image, calcifications, mathematical morphology, analyzes
