DTI in the Cerebal Cortex Correlates with Axon Bundle Organisation: Investigation of Regional Differences in Autism

Background:  Recently there has been increasing interest in applying DTI to the cortex, particularly looking at the primary diffusion direction and how it varies across the cortex. However, to date the structural correlates of this anisotropic diffusion in the cortex have not been examined quantitatively. Therefore, this study looked at the relationship between measures of cortical anisotropy and histological measurements of the radial columnar organisation of the cortex.

Objectives:  To determine the relationship between cortical anisotropy and organisation of axonal bundles and to investigate brain regional differences in autism.

Methods:  Structural and DTI scans were collected post-mortem from 6 ASD brains, 4 control brains and 9 multiple sclerosis brains (a neurological comparison group). The scanning was carried out using a protocol specifically devised for fixed post-mortem tissue as described in Miller et al (2011). Radial cortical anisotropy was analysed using the novel CHIPS software. Following scanning, samples from several regions were removed and sectioned. Sections were either Nissl stained to enable minicolumn measurements to be taken (described in Chance et al (2008)), or stained with Sudan Black to enable measurements of axonal bundles. Image analysis software was used to measure the centre-to-centre spacing of the axon bundles and the cell minicolumns, and the width of the myelinated bundles themselves.

Results:  We found a strong correlation between our measures of radial anisotropy in the cortex and several histological measures of the radial columnar structure. In particular, the measurements of the myelinated bundles showed a good correlation with measures of radial diffusion in the cortex. In addition we found that cortical anisotropy reflects the regional variation in the cortex detected in the histological data.

Conclusions: The present work provides important information on the structural correlates of the observable cortical anisotropy in high-resolution DTI scans and provides validation of the CHIPS technique for investigating cortical correlates of neurological and psychological disorders. Future work to develop this technique for in-vivo use would allow for longitudinal studies of cortical development and alteration in disorders such as autism.