Morphology enabled dipole inversion (MEDI) from a single-angle acquisition: comparison with COSMOS in human brain imaging.

TitleMorphology enabled dipole inversion (MEDI) from a single-angle acquisition: comparison with COSMOS in human brain imaging.
Publication TypeJournal Article
Year of Publication2011
AuthorsLiu T, Liu J, de Rochefort L, Spincemaille P, Khalidov I, Ledoux JRobert, Wang Y
JournalMagn Reson Med
Volume66
Issue3
Pagination777-83
Date Published2011 Sep
ISSN1522-2594
KeywordsAdult, Algorithms, Brain Mapping, Female, Humans, Image Enhancement, Image Processing, Computer-Assisted, Imaging, Three-Dimensional, Linear Models, Magnetic Resonance Imaging, Male, Reproducibility of Results, Sensitivity and Specificity
Abstract

Magnetic susceptibility varies among brain structures and provides insights into the chemical and molecular composition of brain tissues. However, the determination of an arbitrary susceptibility distribution from the measured MR signal phase is a challenging, ill-conditioned inverse problem. Although a previous method named calculation of susceptibility through multiple orientation sampling (COSMOS) has solved this inverse problem both theoretically and experimentally using multiple angle acquisitions, it is often impractical to carry out on human subjects. Recently, the feasibility of calculating the brain susceptibility distribution from a single-angle acquisition was demonstrated using morphology enabled dipole inversion (MEDI). In this study, we further improved the original MEDI method by sparsifying the edges in the quantitative susceptibility map that do not have a corresponding edge in the magnitude image. Quantitative susceptibility maps generated by the improved MEDI were compared qualitatively and quantitatively with those generated by calculation of susceptibility through multiple orientation sampling. The results show a high degree of agreement between MEDI and calculation of susceptibility through multiple orientation sampling, and the practicality of MEDI allows many potential clinical applications.

DOI10.1002/mrm.22816
Alternate JournalMagn Reson Med
PubMed ID21465541
Related Institute: 
MRI Research Institute (MRIRI)

Weill Cornell Medicine
Department of Radiology
525 East 68th Street New York, NY 10065