Identifying global anatomical differences: Deformation-based morphometry

Abstract

The aim of this paper is to illustrate a method for identifying macroscopic anatomical differences among the brains of different populations of subjects. The method involves spatially normalizing the structural MR images of a number of subjects so that they all conform to the same stereotactic space. Multivariate statistics are then applied to the parameters describing the estimated nonlinear deformations that ensue. To illustrate the method, we compared the gross morphometry of male and female subjects. We also assessed brain asymmetry, the effect of handedness, and interactions among these effects.

iLogDemons: A Demons-Based Registration Algorithm for Tracking Incompressible Elastic Biological Tissues

“…In addition, the transformations must be consistent with the properties of the tissue to track, such as elasticity and incompressibility. This is all the more important if the estimated deformations are used to analyse anatomical changes between different time points (Ashburner et al 1998). However, adding these constraints to image registration algorithms are often achieved at the price of computational complexity.…”

A multivariate statistical analysis of the developing human brain in preterm infants

“…It has essentially two goals: (a) to reduce variability due to size, position and orientation of the brain and (b) to reduce variability due to differences in the brain shape. A classification technique can then be categorized into techniques that deal with differences in brain shape (deformation-based morphometry, [2]) and those that deal with differences in the local composition of brain tissue after removing global shape differences (voxel-based morphometry, [1]). While both approaches require warping of images into a standard reference space using either elastic or fluid registration techniques, they differ fundamentally in the way the features are extracted from the aligned images.…”

“…Traditionally, such morphological analysis of brain images has been based either on the definition of regions of interest given some a priori hypothesis or on voxel-wise measurements with little prior knowledge [15,2,17,38,1]. However, as pointed out recently by Lao et al [27], these methodologies have shown practical limitations in their ability to identify new (or subtle) previously unexplored relationships between control and patient populations.…”

Large-scale morphometric analysis of neuroanatomy and neuropathology

“…the use of a brainbased coordinate system is. We use a standardized, or stereotaxic, 3D coordinate space, originally created by Talairach and Tournoux (1988) for neurosurgical applications but now widely accepted in the brain mapping community (Fox et al 1985;Evans et al 1992;Frackowiak 1997) and anatomical variability (Mazziotta et al 1995a, b;Toga and Mazziotta 1996;Ashburner et al 1998). This spatial normalization removes global differences (brain size) between subjects (Collins et al 1994) and provides a framework for the completely automated, 3D analysis across subjects.…”