skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Iterative data refinement of magnetic resonance images

Miscellaneous ·
OSTI ID:6913713

All magnetic resonance (MR) images are blurred as a result of an inherent decaying of nuclear MR signals during data acquisition (DA) due to spatially-varying and object-dependent transverse relaxation (T2). The extent of the blur depends on the distribution of transverse relaxation time of the object and the DA time used in the pulse sequence protocol. Compared to the strength of proton MR signal from a biological organism, sodium signal is inherently weak. A method of improving signal-to-noise ratio in sodium MR imaging is to perform asymmetric sampling of gradient-echo signal so that images with short echo time (2-3 ms) and narrow bandwidth may be acquired. However, a rapid biexponential decay of sodium signal during long DA period, especially due to the presence of fast transverse relaxation (0.7-3 ms) in natural endogenous sodium in tissues, results in fast T2-dependent blurring of reconstructed images. In this dissertation the author considers the problem of correcting for such object-dependent blurs arising from such decay and Fourier transform reconstruction of proton and sodium MR images. Two similar algorithms that correct for such anisotropic blurs in proton and sodium images were developed, implemented, and tested. The first algorithm corrects for mono-exponential T2 distribution and the second algorithm, a natural extension of the first, takes into account the biexponential T2 distribution.. From these algorithms a correction is applied to the raw MR signal. Images reconstructed from such corrected signals yield an improved estimate of T2-weighted spin density distribution. The second algorithm not only corrects for anisotropic blurs in sodium images, but reconstructs decomposed images representing T2-weighted spin density distributions associated with fast and slow T2 of sodium in tissues. Both algorithms tested on mathematical and experimental phantoms show that T2-dependent blurs are reduced.

Research Organization:
Pennsylvania Univ., Philadelphia, PA (United States)
OSTI ID:
6913713
Resource Relation:
Other Information: Thesis (Ph.D.)
Country of Publication:
United States
Language:
English

Similar Records

TU-EF-BRA-00: MR Basics I
Journal Article · Mon Jun 15 00:00:00 EDT 2015 · Medical Physics · OSTI ID:6913713

TU-EF-BRA-01: NMR and Proton Density MRI of the 1D Patient
Journal Article · Mon Jun 15 00:00:00 EDT 2015 · Medical Physics · OSTI ID:6913713

TU-EF-BRA-02: Longitudinal Proton Spin Relaxation and T1-Imaging
Journal Article · Mon Jun 15 00:00:00 EDT 2015 · Medical Physics · OSTI ID:6913713