The accuracy of methods for the detection of mammographic abnormaility is heavily related to breast tissue characteristics. A breast with high tissue density will have reduced sensitivity in terms of detection. Also, breast tissue density is an important indicator of the risk of development of breast cancer. This paper investigates the application of a number of rough set and fuzzy-rough set techniques to mammographic image data. The aim is to attempt to automate the breast tissue classification procedure based on the consensus data of experts. The results of applying the various previously mentioned techniques show that they perform well, achieving high levels of classification accuracy.
|Teitl||Proceedings of the 8th Annual UK Workshop on Computational Intelligence|
|Statws||Cyhoeddwyd - 10 Medi 2008|