Affiliations: [a] Department of Mechanical Engineering, Department of
Radiology, Michigan State University, 2555 Engineering Building, East Lansing,
MI 48824-1226, USA | [b] Department of Electrical & Computer Engineering,
Michigan State University, 2120 Engineering Building, East Lansing, MI
48824-1226, USA
Correspondence:
[*]
Corresponding author: L. Guy Raguin, Department of Mechanical
Engineering, 2555 Engineering Building, East Lansing, 48824-1226, USA. Tel.: +1
517 432 3192; Fax: +1 517 353 1750; E-mail: raguin@msu.edu
Abstract: Diffusion-weighted imaging (DWI) has been shown to provide valuable
structural information in biological fibrous soft tissues, such as the central
and peripheral nervous system and skeletal muscle, based on the acquisition of
a series of images that sample q-space. Existing
q-space sampling schemes have been compared using
several metrics based on diffusion tensor imaging (DTI), which models the
restricted diffusion inside fibres as anisotropic unrestricted diffusion.
Otherwise, sampling schemes have been optimized using model-independent cost
functions. We propose using a model-based framework that connects the
experimental noise, the uncertainty in the estimation of the model parameters,
and the q-space sampling schemes together, to obtain
model-based optimal sampling schemes for the case when the fibre orientation is
known a priori to be within a finite region. We apply and validate this
framework for the recently proposed quantitative analysis of
q-space MRI data (QUAQ) model.
Keywords: Magnetic resonance imaging, diffusion, optimization
