The increase in the number of new breast cancer detections is also increasing the developments of CAD schemes modules in order to provide better accuracy in mammography diagnosis. However, the most of researches in this field have been designed to detect or to classify primary features in mammographic images, as suspect masses and microcalcifications. Asymmetric regions have been considered secondary factors in the breast cancer diagnosis procedure, although they could alert to the existence of a suspect lesion. Therefore, the purpose of this work is to develop a module of a CAD scheme in order to aid the detection and classification - in accordance to the BI-RADS standards - of suspect volume asymmetries regions in digital mammography images. The methodology should be based on intelligent and automated techniques of image processing as well as artificial neural networks. The result of such a work should be inserted as a module of a mammography CAD scheme of our group, in order to enlarge the set of information to be used in each case classification and improving the diagnosis accuracy.
News published in Agência FAPESP Newsletter about the scholarship: