An Information Theory Model for Optimizing Quantitative Magnetic Resonance Imaging Acquisitions
Author ORCID Identifier
Date of Graduation
Doctor of Philosophy (PhD)
David Thomas Alfonso Fuentes, Ph.D.
Ken-Pin Hwang, Ph.D.
James Bankson, Ph.D.
Erik N.K. Cressman, M.D., Ph.D.
James Long, Ph.D.
Quantitative magnetic resonance imaging (qMRI) is a powerful group of imaging techniques with a growing number of clinical applications, including synthetic image generation in post-processing, automatic segmentation, and diagnosis of disease from quantitative parameter values. Currently, acquisition parameter selection is performed empirically for quantitative MRI. Tuning parameters for different scan times, tissues, and resolutions requires some measure of trial and error. There is an opportunity to quantitatively optimize these acquisition parameters in order to maximize image quality and the reliability of the previously mentioned methods which follow image acquisition.
The objective of this work is to introduce and evaluate a quantitative method for selecting parameters that minimize image variability. An information theory framework was developed for this purpose and applied to a 3D-quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) signal model for synthetic MRI. In this framework, mutual information is used to measure the information gained by a measurement as a function of acquisition parameters, quantifying the information content of the acquisition parameters and allowing informed parameter selection.
The information theory framework was tested on synthetic data generated from a representative mathematical phantom, measurements acquired on a qMRI multiparametric imaging standard phantom, and in vivo measurements in a human brain. The application of this information theory framework resulted in successful parameter optimization with respect to mutual information. Both the phantom and in vivo measurements showed that higher mutual information calculated by the model correlated with smaller standard deviation in the reconstructed parametric maps.
With this framework, optimal acquisition parameters can be selected to improve image quality, image repeatability, or scan time. This method could reduce the time and labor necessary to achieve images of the desired quality. Making an informed acquisition parameter selection reduces uncertainty in the imaging output and optimizes information gain within the bounds of clinical constraints.
Information theory, quantitative MRI, mutual information, optimization, image variability