Erin A Paul1, Ana Beatriz Solana2, Jimmy Duong3, Amee M Shah4, Wyman W Lai5, Ek T Tan6, Christopher J Hardy6, Anjali Chelliah4. 1. Division of Pediatric Cardiology, Department of Pediatrics, New York-Presbyterian Morgan Stanley Children's Hospital, Columbia University Medical Center, 3959 Broadway, New York, NY, 10032, USA. eas2188@caa.columbia.edu. 2. GE Healthcare, ASL Europe, Munich, Germany. 3. Department of Biostatistics, Mailman School of Public Health, Columbia University Medical Center, New York, NY, USA. 4. Division of Pediatric Cardiology, Department of Pediatrics, New York-Presbyterian Morgan Stanley Children's Hospital, Columbia University Medical Center, 3959 Broadway, New York, NY, 10032, USA. 5. Department of Pediatric Cardiology, Children's Hospital of Orange County, Orange, CA, USA. 6. GE Global Research, Niskayuna, NY, USA.
Abstract
BACKGROUND: The need for background error correction in phase-contrast flow analysis has historically posed a challenge in cardiac magnetic resonance (MR) imaging. While previous studies have shown that phantom correction improves flow measurements, it impedes scanner workflow. OBJECTIVE: To evaluate the efficacy of self-calibrated non-linear phase-contrast correction on flows in pediatric and congenital cardiac MR compared to phantom correction as the standard. MATERIALS AND METHODS: We retrospectively identified children who had great-vessel phase-contrast and static phantom sequences acquired between January 2015 and June 2015. We applied a novel correction method to each phase-contrast sequence post hoc. Uncorrected, non-linear, and phantom-corrected flows were compared using intraclass correlation. We used paired t-tests to compare how closely non-linear and uncorrected flows approximated phantom-corrected flows. In children without intra- or extracardiac shunts or significant semilunar valvular regurgitation, we used paired t-tests to compare how closely the uncorrected pulmonary-to-systemic flow ratio (Qp:Qs) and non-linear Qp:Qs approximated phantom-corrected Qp:Qs. RESULTS: We included 211 diagnostic-quality phase-contrast sequences (93 aorta, 74 main pulmonary artery [MPA], 21 left pulmonary artery [LPA], 23 right pulmonary artery [RPA]) from 108 children (median age 15 years, interquartile range 11-18 years). Intraclass correlation showed strong agreement between non-linear and phantom-corrected flow measurements but also between uncorrected and phantom-corrected flow measurements. Non-linear flow measurements did not more closely approximate phantom-corrected measurements than did uncorrected measurements for any vessel. In 39 children without significant shunting or regurgitation, mean non-linear Qp:Qs (1.07; 95% confidence interval [CI] = 1.01, 1.13) was no closer than mean uncorrected Qp:Qs (1.06; 95% CI = 1.00, 1.13) to mean phantom-corrected Qp:Qs (1.02; 95% CI = 0.98, 1.06). CONCLUSION: Despite strong agreement between self-calibrated non-linear and phantom correction, cardiac flows and shunt calculations with non-linear correction were no closer to phantom-corrected measurements than those without background correction. However, phantom-corrected flows also demonstrated minimal differences from uncorrected flows. These findings suggest that in the current era, more accurate phase-contrast flow measurements might limit the need for background correction. Further investigation of the clinical impact and optimal methods of background correction in the pediatric and congenital cardiac population is needed.
BACKGROUND: The need for background error correction in phase-contrast flow analysis has historically posed a challenge in cardiac magnetic resonance (MR) imaging. While previous studies have shown that phantom correction improves flow measurements, it impedes scanner workflow. OBJECTIVE: To evaluate the efficacy of self-calibrated non-linear phase-contrast correction on flows in pediatric and congenital cardiac MR compared to phantom correction as the standard. MATERIALS AND METHODS: We retrospectively identified children who had great-vessel phase-contrast and static phantom sequences acquired between January 2015 and June 2015. We applied a novel correction method to each phase-contrast sequence post hoc. Uncorrected, non-linear, and phantom-corrected flows were compared using intraclass correlation. We used paired t-tests to compare how closely non-linear and uncorrected flows approximated phantom-corrected flows. In children without intra- or extracardiac shunts or significant semilunar valvular regurgitation, we used paired t-tests to compare how closely the uncorrected pulmonary-to-systemic flow ratio (Qp:Qs) and non-linear Qp:Qs approximated phantom-corrected Qp:Qs. RESULTS: We included 211 diagnostic-quality phase-contrast sequences (93 aorta, 74 main pulmonary artery [MPA], 21 left pulmonary artery [LPA], 23 right pulmonary artery [RPA]) from 108 children (median age 15 years, interquartile range 11-18 years). Intraclass correlation showed strong agreement between non-linear and phantom-corrected flow measurements but also between uncorrected and phantom-corrected flow measurements. Non-linear flow measurements did not more closely approximate phantom-corrected measurements than did uncorrected measurements for any vessel. In 39 children without significant shunting or regurgitation, mean non-linear Qp:Qs (1.07; 95% confidence interval [CI] = 1.01, 1.13) was no closer than mean uncorrected Qp:Qs (1.06; 95% CI = 1.00, 1.13) to mean phantom-corrected Qp:Qs (1.02; 95% CI = 0.98, 1.06). CONCLUSION: Despite strong agreement between self-calibrated non-linear and phantom correction, cardiac flows and shunt calculations with non-linear correction were no closer to phantom-corrected measurements than those without background correction. However, phantom-corrected flows also demonstrated minimal differences from uncorrected flows. These findings suggest that in the current era, more accurate phase-contrast flow measurements might limit the need for background correction. Further investigation of the clinical impact and optimal methods of background correction in the pediatric and congenital cardiac population is needed.
Authors: Daniel Giese; Maximilian Haeberlin; Christoph Barmet; Klaas P Pruessmann; Tobias Schaeffter; Sebastian Kozerke Journal: Magn Reson Med Date: 2011-08-08 Impact factor: 4.668
Authors: P G Walker; G B Cranney; M B Scheidegger; G Waseleski; G M Pohost; A P Yoganathan Journal: J Magn Reson Imaging Date: 1993 May-Jun Impact factor: 4.813
Authors: Cynthia K Rigsby; Nicholas Hilpipre; Gary R McNeal; Gang Zhang; Emma E Boylan; Andrada R Popescu; Grace Choi; Andreas Greiser; Jie Deng Journal: Pediatr Radiol Date: 2013-12-05
Authors: Peter D Gatehouse; Marijn P Rolf; Martin J Graves; Mark Bm Hofman; John Totman; Beat Werner; Rebecca A Quest; Yingmin Liu; Jochen von Spiczak; Matthias Dieringer; David N Firmin; Albert van Rossum; Massimo Lombardi; Juerg Schwitter; Jeanette Schulz-Menger; Philip J Kilner Journal: J Cardiovasc Magn Reson Date: 2010-01-14 Impact factor: 5.364
Authors: Krishna S Nayak; Jon-Fredrik Nielsen; Matt A Bernstein; Michael Markl; Peter D Gatehouse; Rene M Botnar; David Saloner; Christine Lorenz; Han Wen; Bob S Hu; Frederick H Epstein; John N Oshinski; Subha V Raman Journal: J Cardiovasc Magn Reson Date: 2015-08-09 Impact factor: 5.364