Premium
Fetal Flow Quantification in Great Vessels Using Motion‐Corrected Radial Phase Contrast MRI : Comparison With Cartesian
Author(s) -
Goolaub Datta Singh,
Xu Jiawei,
Schrauben Eric,
Sun Liqun,
Roy Christopher W.,
Marini Davide,
Seed Mike,
Macgowan Christopher K.
Publication year - 2021
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.27334
Subject(s) - cartesian coordinate system , imaging phantom , contrast (vision) , nuclear medicine , physics , mathematics , medicine , artificial intelligence , computer science , geometry
Background Phase contrast MRI in the great vessels is a potential clinical tool for managing fetal pathologies. One challenge is the uncontrollable fetal motion, potentially corrupting flow quantifications. Purpose To demonstrate improvements in fetal blood flow quantification in great vessels using retrospectively motion‐corrected golden‐angle radial phase contrast MRI relative to Cartesian phase contrast MRI. Study Type Method comparison. Phantom/Subjects Computer simulation. Seventeen pregnant volunteers. Field Strength/Sequence 1.5T and 3T . Cartesian and golden‐angle radial phase contrast MRI . Assessment Through computer simulations, radial (with and without retrospective motion correction) and Cartesian phase contrast MRI were compared using flow deviations. in vivo Cartesian and radial phase contrast MRI measurements and reconstruction qualities were compared in pregnancies. Cartesian data were reconstructed into gated reconstructions (CINEs) after cardiac gating with metric optimized gating (MOG). For radial data, real‐time reconstructions were performed for motion correction and MOG followed by CINE reconstructions. Statistical Tests Wilcoxon signed‐rank test. Linear regression. Bland–Altman plots. Student's t ‐test. Results Simulations showed significant improvements ( P < 0.05) in flow accuracy and reconstruction quality with motion correction ([mean/peak] flow errors with ±5 mm motion corruption: Cartesian [35 ± 1/115 ± 7] mL/s, motion uncorrected radial [25 ± 1/75 ± 2] mL/s and motion‐corrected radial [1.0 ± 0.5/−5 ± 1] mL/s). in vivo Cartesian reconstructions without motion correction had lower quality than the motion‐corrected radial reconstructions ( P < 0.05). Across all fetal mean flow measurements, the bias [limits of agreement] between the two measurements were −0.2 [−76, 75] mL/min/kg, while the linear regression coefficients were (M radial = 0.81 × M Cartesian + 29.8 [mL/min/kg], r 2 = 0.67). The corresponding measures for the peak fetal flows were −23 [−214, 167] mL/min/kg and ( P radial = 0.95 × P Cartesian –1.2 [mL/min/kg], r 2 = 0.80). Cartesian reconstructions of low quality showed significantly higher estimated mean and peak ( P < 0.05) flows than the corresponding radial reconstructions. Data Conclusion Simulations showed that radial phase contrast MRI with motion compensation improved flow accuracy. For fetal measurements, motion‐corrected radial reconstructions showed better image quality than, and different flow values from, Cartesian reconstructions. Level of Evidence 1. Technical Efficacy Stage 1. J. MAGN. RESON. IMAGING 2021;53:540–551.