Purpose

Variation of grey matter (GM) myelination in the human brain has been used in histology and more recently in vivo MRI to delineate and classify areas with different myeloarchitecture, identifying heavily myelinated cortical areas (primary area) such as V1 and V5 area in the visual cortex. The white matter (WM) tracts connecting these primary area are believed to be myelinated more heavily during the development of the brain, while the corresponding cortical ribbon is thinning¹. The aim of this study is to estimate the correlation between myelination in the cortical ribbon and the underlying subcortical WM fibres in healthy adults.

Methods

58 healthy volunteers (39 ± 12 years old, 27 male) gave written informed consent to participate in this study approved by the Ethics Committee. MRI data were collected using a Philips Achieva 7T system. A T₁-weighted image, based on the Phase Sensitive Inversion Recovery sequence (PSIR; FOV = 240x216x160mm³, 0.8mm isotropic resolution, TI₁/TI₂=780ms/1600ms), was acquired and used for segmentation, T₁ mapping and MEG coregistration. MT data was acquired using an MT-TFE sequence (saturation train: 20 gaussian-windowed sinc pulses, BW=200Hz, 17 saturation frequency from -5kHz to 5kHz, B_1rms values of 0.33, 0.65 and 1.09μT; acquisition: TE/TR/FA = 2.7ms/5.8ms/8°, FOV = 192x192x60 mm³, 1.5mm isotropic image resolution, low-high k-space acquisition and a SENSE factor (RL) of 2). Three z-spectra were acquired for each subject for the three saturation powers, and quantitative MT data was derived by fitting the three z-spectra to a Look-Up-Table database of simulated spectra, based on the Bloch-McConnell equations². Automatic segmentation using freesurfer³, followed by iterative manual correction was performed for each individual. Cortical GM and adjacent subcortical WM was classified using the Desikan-Killiany atlas⁴, using a 5 voxels metric to assign WM a similar label to the closest GM region, as shown in Figure 1. Mean MT value was extracted for each region, for each participant. The modal value for each individual’s MT data was regressed from that individual’s regional values. Pearson correlation, measured across subjects, was used to quantify the relationship between MT values measured in Desikan-Killiany region pairs. A permutation test statistically assessed the relationship between each of the cortical regions (N=10000 permutations), and the resulting p-values were corrected for multiple comparison using the Benjamini and Hochberg procedure. Pearson correlation of MT with cortical thickness extracted via freesurfer, as well as pseudo T₁ obtained from the PSIR were also tested. The data was additionally regressed for B₁ effect, and regions with missing voxels were removed from further analysis.

Results

The MT in 19 GM regions per hemisphere were compared to the corresponding subcortical WM regions, and a strong correlation was found throughout the brain (mean R value of 0.76, figure 2). All regions had significant difference (p<0.05) between the permutation test and the original pair (figure 3 b and a), which persisted following multiple comparison correction. After regression of the B₁ from the MT data, the significant difference remained in all regions. The average correlation remained strongly significant, with a mean R value of 0.72 between GM regions and adjacent WM regions, with variation from 0.61 in the parstriangularis right hemisphere, up to 0.88 in the precuneus left hemisphere, as represented in figure 4. No correlation could be seen between cortical thickness and MT either in WM or in GM (mean R value of respectively 0.01 and 0.04), while a moderate correlation (mean R value of 0.56) was found for the pseudo T₁ between GM and subcortical WM.

Discussion

Myelin content is highly correlated between the GM and underlying subcortical WM for the GM regions studied here. The cortical GM microstructure is highly specialised to the function needed, and myelination is a strong part of this specialization process. WM axonal myelination develop to support the GM activity and connectivity. WM myelination occurs throughout early childhood, with a recent study showing myelin water fraction increasing similarly in white and grey matter regions¹. It would be interesting to repeat this analysis through childhood to observe the development of cortical and white matter myelination. The present study shows that MT is sensitive enough to measure variation of myelin throughout the brain in vivo, and present a new insight into regional specialisation of the brain.

Conclusion

Myelination can be measured using quantitative MT. The results shows that GM myeloarchitectonic is reflected in the underlying WM in healthy adults.

Acknowledgements

This work was funded by Medical Research Council (MRC) UK Partnership Grant, MR/K005464/1, MRC New Investigator Research Grant (MR/M006301/1), and MRC Doctoral Training Grant, MR/K501086/1.