Diffusion tensor imaging has been widely used in studying rodent models of white matter diseases. In this study, we examined the differences between in vivo and ex vivo fractional anisotropy and diffusivity measurements in the mouse cuprizone model. In the control mouse corpus callosum, ex vivo diffusivities were significantly lower than in vivo measurements, but ex vivo fractional anisotropy values were not significantly different from in vivo fractional anisotropy values. With cuprizone induced demyelination and accompanying pathology in the corpus callosum, changes in in vivo and ex vivo fractional anisotropy and diffusivity measurements were not always in agreement. Our results suggest that ex vivo λ(⟂) was a more reliable indicator of white matter demyelination than in vivo λ(⟂) and in vivo λ(‖) was a more reliable indicator of axonal injury than ex vivo λ(‖) in this model. When comparing in vivo and ex vivo diffusion tensor imaging results of axon and myelin pathology in the rodent models, potential changes in tissue microstructures associated with perfusion fixation should be considered.
Copyright © 2011 Wiley Periodicals, Inc.