Apparent diffusion coefficients from high angular resolution diffusion imaging: Estimation and applications

Abstract

High angular resolution diffusion imaging has recently been of great interest in characterizing non-Gaussian diffusion processes. One important goal is to obtain more accurate fits of the apparent diffusion processes in these non-Gaussian regions, thus overcoming the limitations of classical diffusion tensor imaging. This paper presents an extensive study of highorder models for apparent diffusion coefficient estimation and illustrates some of their applications. Using a meaningful modified spherical harmonics basis to capture the physical constraints of the problem, a new regularization algorithm is proposed. The new smoothing term is based on the LaplaceBeltrami operator and its closed form implementation is used in the fitting procedure. Next, the linear transformation between the coefficients of a spherical harmonic series of order ഞ and independent elements of a rank-ഞ high-order diffusion tensor is explicitly derived. This relation allows comparison of the stateof-the-art anisotropy measures computed from spherical harmonics and tensor coefficients. Published results are reproduced accurately and it is also possible to recover voxels with isotropic, single fiber anisotropic, and multiple fiber anisotropic diffusion. Validation is performed on apparent diffusion coefficients from synthetic data, from a biological phantom, and from a human brain dataset. For the past decade, there has been growing interest in diffusion magnetic resonance imaging (MRI) to understand functional coupling between cortical regions of the brain, for characterization of neurodegenerative diseases, for surgical planning, and for other medical applications. Diffusion MRI is the only noninvasive tool to obtain information about the neural architecture in vivo. It is based on the Brownian motion of water molecules in normal tissues and the observation that molecules tend to diffuse along fibers when contained in fiber bundles (1,2). Using classic diffusion tensor imaging (DTI), several methods have been developed to segment and track white matter fibers in the human brain (3-8). The common way to analyze the data is to fit it to a second-order tensor, which corresponds to the probability distribution of a given water molecule moving by a certain amount during some fixed elapsed time. By diagonalization, the surface corresponding to the diffusion tensor is an ellipsoid with its long axis aligned with the fiber orientation. However, the theoretical basis for this model assumes that the underlying diffusion process is Gaussian. While this approximation is adequate for voxels in which there is only a single fiber orientation (or none). it breaks down for voxels in which there is more complicated internal structure, as seen in Fig. 1, an example of two fibers crossing. This is an important limitation, since the resolution of DTI acquisition is between 1 and 3 mm 3 , while the physical diameter of fibers can be less than 1 m and up to 30 m (9). From anisotropy measure maps, we know that many voxels in diffusion MRI volumes potentially have m...

Beyond fractional anisotropy: Extraction of bundle-specific structural metrics from crossing fiber models

“…The particular SH base functions used here have been proposed by Descoteaux et al (2006). For rank m and order l they are defined as: where P ‹ X is the Legendre polynomial of rank m and order l. As the measured diffusion signal within each voxel is naturally positive and symmetric with respect to the origin, only spherical harmonic base functions of even rank have to be considered to represent the fODF.…”

DiffusionKit: A light one-stop solution for diffusion MRI data analysis

“…Generally, under a diffusion-weighted spin-echo EPI pulse sequence, the propagation of water molecules will be hindered by the white matter microstructures, and the detected DWI attenuated signal ( ) is expressed as (Cheng et al, 2013;Descoteaux et al, 2006) …”

Efficient and robust computation of PDF features from diffusion MR signal

“…The λ l and λ n set the angular and radial weights of the regularization. Similarly to [57], angular regularization is equivalent to the application of a smoothing filter in the angular q-space domain. This comes from the duality between the convolution in the q-space and the multiplication in the spherical harmonics domain [58].…”