Neural Circuits and Systems Neuroscience
Author: Ines Samengo | email: ines.samengo@gmail.com
Bautista Arenaza 1°, Mariana Vallejo Azar 2°, Bautista Elizalde Acevedo 3°, Sergio Lindenbaum 4°, Lucía Alba Ferrara 2°, Mariana Bendersky 2°, Paula González 2°, Sebastián Risau Gusmán 1°, Inés Samengo 1°
1° Department of Medical Physics, Centro Atómico Bariloche, CONICET.
2° Estudios en Neurociencias y Sistemas Complejos, ENyS (Consejo Nacional de Investigaciones Científicas y Técnicas, Universidad Nacional Arturo Jauretche, Hospital El Cruce)
3° Instituto de Investigaciones en Medicina Traslacional (IIMT), CONICET-Universidad Austral.
4° Instituto de Tecnologías Nucleares para la Salud, Bariloche, Río Negro.
MRI scans provide highly detailed information about human neuroanatomical data. Here we model the population statistics of the geometrical properties of different cortical areas of a sample of 193 South American healthy adults (both sexes, aged 18-60) using a hierarchical Bayesian model. We find that, at the population level, areas and thicknesses of different cortical parcels are largely independent from each other, suggesting that the factors that determine the area of a region are independent from those shaping the thickness. Both types of measures exhibit strong correlations with the homologous contralateral brain region, except for prefrontal regions identified with language processing. Intra-hemispheric correlations are shown to be smaller in size than inter-hemispheric ones. The fact that they are always positive suggests that there is no evident competition for space between different regions. A segmentation of the correlation graph in communities reveals that brain regions whose areas or thicknesses are correlated tend to be anatomically contiguous. This result suggests that the events that determine the size of a given brain area affect also its neighbours, as expected, for example, during early brain development stages.