Simplified gyral pattern in severe developmental microcephalies? New insights from allometric modeling for spatial and spectral analysis of gyrification.

The strong positive-allometric relationship between brain size, cortical extension and gyrification complexity, recently highlighted in the general population, could be modified by brain developmental disorders. Indeed, in case of brain growth insufficiency, the pathophysiological relevance of the "simplified gyral pattern" phenotype is strongly disputed since almost no genotype-phenotype correlations have been found in primary microcephalies. Using surface scaling analysis and newly-developed spectral analysis of gyrification (Spangy), we tested whether the gyral simplification in groups of severe microcephalies related to ASPM, PQBP1 or fetal-alcohol-syndrome could be fully explained by brain size reduction according to the allometric scaling law established in typically-developing control groups, or whether an additional disease effect was to be suspected. We found the surface area reductions to be fully explained by scaling effect, leading to predictable folding intensities measured by gyrification indices. As for folding pattern assessed by spectral analysis, scaling effect also accounted for the majority of the variations, but an additional negative or positive disease effect was found in the case of ASPM and PQBP1-linked microcephalies, respectively. Our results point out the necessity of taking allometric scaling into account when studying the gyrification variability in pathological conditions. They also show that the quantitative analysis of gyrification complexity through spectral analysis can enable distinguishing between even (predictable, non-specific) and uneven (unpredictable, maybe disease-specific) gyral simplifications.

Model-driven parameterization of the cortical surface for localization and inter-subject matching.

In this paper we present a generic and organized model of cortical folding, and a way to implement this model on any given cortical surface. This results in a model-driven parameterization, providing an anatomically meaningful coordinate system for cortical localization, and implicitly defining inter-subject surface matching without any deformation of surfaces. We present our cortical folding model and show how it naturally defines a parameterization of the cortex. The mapping of the model to any given cortical surface is detailed, leading to an anatomically invariant coordinate system. The process is evaluated on real data in terms of both anatomical and functional localization, and shows improved performance compared to a traditional volume-based normalization. It is fully automatic and available with the BrainVISA software platform.

Projection of fMRI data onto the cortical surface using anatomically-informed convolution kernels.

As surface-based data analysis offer an attractive approach for intersubject matching and comparison, the projection of voxel-based 3D volumes onto the cortical surface is an essential problem. We present here a method that aims at producing representations of functional brain data on the cortical surface from functional MRI volumes. Such representations are for instance required for subsequent cortical-based functional analysis. We propose a projection technique based on the definition, around each node of the gray/white matter interface mesh, of convolution kernels whose shape and distribution rely on the geometry of the local anatomy. For one anatomy, a set of convolution kernels is computed that can be used to project any functional data registered with this anatomy. Therefore resulting in anatomically-informed projections of data onto the cortical surface, this kernel-based approach offers better sensitivity, specificity than other classical methods and robustness to misregistration errors. Influences of mesh and volumes spatial resolutions were also estimated for various projection techniques, using simulated functional maps.