Premium
Quantitative susceptibility mapping across two clinical field strengths: Contrast‐to‐noise ratio enhancement at 1.5T
Author(s) -
Ippoliti Matteo,
Adams Lisa C.,
Winfried Brenner,
Hamm Bernd,
Spincemaille Pascal,
Wang Yi,
Makowski Marcus R.
Publication year - 2018
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.26045
Subject(s) - reproducibility , wilcoxon signed rank test , quantitative susceptibility mapping , repeatability , nuclear medicine , rank correlation , mathematics , correlation , voxel , nuclear magnetic resonance , linear regression , medicine , magnetic resonance imaging , biomedical engineering , statistics , physics , radiology , mann–whitney u test , geometry
Background Quantitative susceptibility mapping (QSM) is an MRI postprocessing technique that allows quantification of the spatial distribution of tissue magnetic susceptibility in vivo. Contributing sources include iron, blood products, calcium, myelin, and lipid content. Purpose To evaluate the reproducibility and consistency of QSM across clinical field strengths of 1.5T and 3T and to optimize the contrast‐to‐noise ratio (CNR) at 1.5T through bandwidth tuning. Study Type Prospective. Subjects Sixteen healthy volunteers (10 men, 6 women; age range 24–37; mean age 27.8 ± 3.2 years). Field Strengths/Sequences 1.5T and 3T systems from the same vendor. Four spoiled gradient echo (SPGR) sequences were designed with different acquisition bandwidths. Assessment QSM reconstruction was achieved through a nonlinear morphology‐enabled dipole inversion (MEDI) algorithm employing L1 regularization. CNR was calculated in seven regions of interest (ROIs), while reproducibility and consistency of QSM measurements were evaluated through voxel‐based and region‐specific linear correlation analyses and Bland–Altman plots. Statistical Tests Interclass correlation, Wilcoxon rank sum test, linear regression analysis, Bland–Altman analysis, Welch's t ‐test. Results CNR analysis showed a statistically significant ( P < 0.05) increase in four out of seven ROIs for the lowest bandwidth employed with respect to the highest (25.18% increase in CNR of caudate nucleus). All sequences reported an excellent correlation across field strength and bandwidth variation ( R ≥ 0.96, widest limits of agreement from –18.7 to 25.8 ppb) in the ROI‐based analysis, while the correlation was found to be good for the voxel‐based analysis of averaged maps ( R ≥ 0.90, widest limits of agreement from –9.3 to 9.1 ppb). Data Conclusion CNR of QSM images reconstructed from 1.5T acquisitions can be enhanced through bandwidth tuning. MEDI‐based QSM reconstruction demonstrated to be reproducible and consistent both across field strengths (1.5T and 3T) and bandwidth variation. Level of Evidence: 1 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2018;47:1410–1420.