WSEAS Transactions on Biology and Biomedicine
Print ISSN: 1109-9518, E-ISSN: 2224-2902
Volume 11, 2014
Model Based Non-Rigid Registration Framework For High Dynamic Range Mammography
Authors: , ,
Abstract: Mammography screening is the most prominent method for finding breast cancer at an early stage. The acquired mammograms are High Dynamic Range images having a 12 bit gray scale resolution. When viewed by a radiologist, a single image must be examined several times, each time focusing on a different intensity range. So this paper proposes the Computer aided approach for enhancing standard lesions in digital mammograms, where we have developed a Model based Non-rigid Registration Framework (MBNRF) for High Dynamic Range mammogram image enhancement in a fully automatic way. The proposed system considers two views of mammogram images such as MLO (Mediolateral Oblique View) and CC (Craniocaudal) for processing and consists of two parts: 1) Preliminary processing operations involves in use of Contrast limited Adaptive Histogram Equalization technique to remove noise and intensity in homogeneities and 2) Registering the CC (Craniocaudal) model with MLO (Mediolateral Oblique View) mammogram view image by entropy based registration. The algorithm’s performance has been tested on few mammography images in collaboration with radiologists.
Search Articles
Pages: 126-132
WSEAS Transactions on Biology and Biomedicine, ISSN / E-ISSN: 1109-9518 / 2224-2902, Volume 11, 2014, Art. #17