Statistical analysis of morphometric measures based on labeled cortical distance maps

Abstract

Shape differences in cortical structures in the brain can be associated with various neuropsychiatric and neuro-developmental diseases or disorders. Labeled Cortical Distance Map (LCDM) can be a powerful tool to quantize such differences in shapes derived from magnetic resonance images (MRI). This article investigates some aspects of LCDM distances in relation to morphometry. Simple morphometric measures based on LCDM indicate some aspect of the shape or size of the tissue in question. The length of the LCDM distance vector provides the number of voxels and thus volume of the tissue. The median, mode, range, and variance of LCDM distances and volume of the tissue are all suggestive of size, thickness, and shape differences. Statistical tests are employed to detect left-right asymmetry, group differences, and stochastic ordering (cdf differences) of these LCDM-based variables. We perform LCDM analysis of gray matter in ventral medial prefrontal cortices (VMPFCs) obtained from a neuro-imaging study of major depressive disorder (MDD), high risk, and control twin subjects. We find significant evidence that VMPFCs with MDD exhibit significant morphometric left-right asymmetry compared to those in high risk and control subjects. The method is also valid for analysis of morphometric measures of other organs or tissues and distances similar to LCDM distances.