Predicting degree of myelination based on diffusion tensor imagining of canines with mucopolysaccharidosis type I

Abstract

Objective

This study examined the effect of mucopolysaccharidosis (MPS) type 1 on diffusion tensor imaging (DTI) metrics in the canine brain. We hypothesized 1) white matter regions in the MPS brain would show decreased fractional anisotropy (FA) and increased radial diffusivity (RD) compared to the same regions in normal brain, 2) compared to FA, RD would more closely correlate with myelin density and fiber coherence, and 3) DTI and histological data from the normal brain could be used to accurately predict degree of myelination in the MPS brain using DTI metrics.

Methods

We performed DTI imaging on one normal canine brain and two MPS brains on a 7T MR scanner and generated FA and RD maps. Brains were sectioned and stained with a gold chloride stain for myelin to obtain myelin optical density and fiber coherence values. The three brains were compared using the DTI and histology metrics.

Results

Most measured regions in one MPS brain and all measured regions in the other MPS brain exhibited decreased FA, increased RD, and decreased myelin density in white matter. FA and RD significantly correlated with myelin density in the normal brain but failed to reach significance in either MPS brain. A predictive model using FA but not RD was able to accurately predict degree of myelination in one MPS brain.

Conclusion

Dysmyelination in the MPS brain results in decreased FA and increased RD. However, in the small sample, FA and RD were values not significantly correlated with myelination in either MPS brain.