Robust anatomical landmark detection for MR brain image registration

Abstract

Correspondence matching between MR brain images is often challenging due to large inter-subject structural variability. In this paper, we propose a novel landmark detection method for robust establishment of correspondences between subjects. Specifically, we first annotate distinctive landmarks in the training images. Then, we use regression forest to simultaneously learn (1) the optimal set of features to best characterize each landmark and (2) the non-linear mappings from local patch appearances of image points to their displacements towards each landmark. The learned regression forests are used as landmark detectors to predict the locations of these landmarks in new images. Since landmark detection is performed in the entire image domain, our method can cope with large anatomical variations among subjects. We evaluated our method by applying it to MR brain image registration. Experimental results indicate that by combining our method with existing registration method, obvious improvement in registration accuracy can be achieved.

Fig. 1

The six manually annotated landmarks on the corners/edges of the ventricular region in one training image. The white cross marks denote the landmark locations.

Fig. 2

Registration results on 5 testing images that are substantially different in appearance from the fixed image. Row 1: The fixed image and the original moving images. Row 2: Registration results given by HAMMER alone. Row 3: Registration results given by landmark-based initialization. Row 4: Registration results given by further refinement using HAMMER.