Unbiased Atlas Formation Via Large Deformations Metric Mapping

Abstract

Abstract. The construction of population atlases is a key issue in medical image analysis, and particularly in brain mapping. Large sets of images are mapped into a common coordinate system to study intrapopulation variability and inter-population differences, to provide voxelwise mapping of functional sites, and to facilitate tissue and object segmentation via registration of anatomical labels. We formulate the unbiased atlas construction problem as a Fréchet mean estimation in the space of diffeomorphisms via large deformations metric mapping. A novel method for computing constant speed velocity fields and an analysis of atlas stability and robustness using entropy are presented. We address the question: how many images are required to build a stable brain atlas?

Atlas-based delineation of lymph node levels in head and neck computed tomography images

“…It has the advantage of being faster and simpler than the one proposed by Lorenzen et al [21] and is su cient for our images. This method is based on an iterative scheme to build an unbiased mean image from the image database.…”

“…They iterate on two steps: the registration of the images on the reference and the application of the inverse mean transformation to the mean image. Guimond et al [19] have shown that this approach, extended by [20,21] to transformations containing large deformations, is not dependent on the choice of the reference image. Recently, based on this principle, a method has been investigated to generate directly a mean symmetric atlas from a database of images [22].…”

A mean three-dimensional atlas of the human thalamus: Generation from multiple histological data

“…Evans et al (1993) averaged 305 MRIs to construct such a mean atlas. Their work was improved through the use of non-linear registration algorithms (Seghers et al, 2004;Lorenzen et al, 2005;Christensen et al, 2006). Building an atlas from a population also allows to capture the anatomical variability of the structures of interest (Pohl et al, 2004;Styner et al, 2003).…”

Multiscale Unbiased Diffeomorphic Atlas Construction on Multi-GPUs

“…To evaluate the robustness and stability of the atlases, we use the random permutation test proposed by Lorenzen et al [10]. The method is capable of estimating the minimum number of inputs required to construct a stable atlas by analyzing mean entropy and the variance of the average template.…”