Improving microcalcification detection using local texture features

OData support
Supervisor:
Dr. Pataki Béla
Department of Measurement and Information Systems

In modern societies about 10% of women’s gets some type of breast cancer, thanks to the modern medical most of them are curable. It’s known that the most important in the healing, the prevention, early detection and filtering. Nowadays there are a lot of filter examination for women’s, but this means about one million x-ray pictures yearly only in Hungary. To evaluate so many pictures we need a lot of well qualified doctors, it would be advantageous to build an application which can help in pre filtering and help to make the right decisions, with the raising of the suspect parts.

One from the most important properties of the breast cancer is the micro calcification, but there are several problems in the calcification detection, the dispersion of these textures and breast types are high, the distribution of the fat and the connective tissue can show big differences, and this results varied intensity and texture in the x-ray images. And sometimes in different parts of one breast shows big differences. In this thesis we try to develop an application which can detect calcifications in x-ray pictures, based on local texture properties.

Downloads

Please sign in to download the files of this thesis.