Purpose Neurodegenerative diseases such as Alzheimer’s disease cause changes and disruption to cortical microstructure and architecture. Diffusion MRI (dMRI) could potentially be sensitive to such changes. There is a growing interest in modeling of human cortical areas using a combination of quantitative MRI and 3D microscopy. The purpose of this study was to quantitatively characterize the cytoarchitecture of human cortical tissue from 3D fluorescence microscopy to simulate diffusion MRI (dMRI) signal in the cortex to better understand its diffusion signal characteristics.Methods Diffusion of water molecules and dMRI signal were simulated by an indirect geometry based method and a direct voxel based method in microstructural details extracted from microscopy of cortex. Additionally, residence times of diffusing spins inside voxel volumes were considered to set effective resolution limits. Mean diffusivity (MD) and kurtosis (MK) were calculated for variable cell and neurite densities, sizes and diffusion times under realistic values for permeability and free diffusion.Results Both simulation methods could efficiently and accurately simulate dMRI signals with fractional anisotropy, diffusion coefficient and kurtosis in agreement with previous reports. Simulated MD and MK showed changes with increasing diffusion times specific to cortical cell density and sizes, with MK showing the highest sensitivity. Intra-voxel residence times with increasing diffusion times showed that the effective dMRI resolution approaches the thickness of cortical layers.Conclusions Monte Carlo simulations based on 3D microscopy data enable estimating changes in MD and MK over diffusion times and are sensitive to cortical cytoarchitecture and its possible changes in neurodegenerative disease. When considering layer-specific cortical dMRI, effective resolution due to residence times is an important concern.
|Publisher||Cold Spring Harbor Laboratory - bioRxiv|
|Publication status||Published - 4 May 2019|