Detecting mutually-salient landmark pairs with MRF regularization

• Published in: 2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro pp. 400 - 403
• Main Authors: Ou, Y, Besbes, A, Bilello, M, Mansour, M, Davatzikos, C, Paragios, N
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.04.2010
• Subjects: Biomedical imaging; Biomedical measurements; Data mining; Image analysis; Image registration; Landmark Matching; MRF; Mutual information; Mutual-Saliency; Optimization methods; Radiology
• ISBN: 9781424441259; 1424441250
• ISSN: 1945-7928; 1945-8452
• DOI: 10.1109/ISBI.2010.5490324

In this paper, we present a framework for extracting mutually-salient landmark pairs for registration. Traditional methods detect landmarks one-by-one and separately in two images. Therefore, the detected landmarks might inherit low dis-criminability and are not necessarily good for matching. In contrast, our method detects landmarks pair-by-pair across images, and those pairs are required to be mutually-salient, i.e., uniquely corresponding to each other. The second merit of our framework is that, instead of finding individually optimal correspondence, which is a local approach and could cause self-intersection of the resultant deformation, our framework adopts a Markov-random-field (MRF)-based spatial arrangement to select the globally optimal landmark pairs. In this way, the geometric consistency of the correspondences is maintained and the resultant deformations are relatively smooth and topology-preserving. Promising experimental validation through a radiologist's evaluation of the established correspondences is presented.