Nonparametric tests of structure for high angular resolution diffusion imaging in Q-space - Statistics > ApplicationsReport as inadecuate




Nonparametric tests of structure for high angular resolution diffusion imaging in Q-space - Statistics > Applications - Download this document for free, or read online. Document in PDF available to download.

Abstract: High angular resolution diffusion imaging data is the observed characteristicfunction for the local diffusion of water molecules in tissue. This data isused to infer structural information in brain imaging. Nonparametric scalarmeasures are proposed to summarize such data, and to locally characterizespatial features of the diffusion probability density function PDF, relyingon the geometry of the characteristic function. Summary statistics are definedso that their distributions are, to first-order, both independent of nuisanceparameters and also analytically tractable. The dominant direction of thediffusion at a spatial location voxel is determined, and a new set of axesare introduced in Fourier space. Variation quantified in these axes determinesthe local spatial properties of the diffusion density. Nonparametric hypothesistests for determining whether the diffusion is unimodal, isotropic ormulti-modal are proposed. More subtle characteristics of white-mattermicrostructure, such as the degree of anisotropy of the PDF and symmetrycompared with a variety of asymmetric PDF alternatives, may be ascertaineddirectly in the Fourier domain without parametric assumptions on the form ofthe diffusion PDF. We simulate a set of diffusion processes and characterizetheir local properties using the newly introduced summaries. We show howcomplex white-matter structures across multiple voxels exhibit clearellipsoidal and asymmetric structure in simulation, and assess the performanceof the statistics in clinically-acquired magnetic resonance imaging data.



Author: Sofia C. Olhede, Brandon Whitcher

Source: https://arxiv.org/







Related documents