DESENVOLVIMENTO DE EXTRATORES DE CARACTERÍSTICAS PARA CLASSIFICAR TUMORES DE MAMA

The breast cancer is one of the major reasons of de ath among women. Computer aided diagnostic systems must be able to automatically id entify existent features in mammograms. These features must discriminate benign and malign tumors, or retrieve images correctly. The success of computer aided diagnostic system is stro ngly associated with the features that represent the radiological finding in a mammogram. This work presents two shape feature extractors (Complexity Index and Elliptic Variance) , that extract information from the lesion. The