Purpose: To develop and evaluate a cardiac phase-resolved myocardial T1 mapping sequence. Methods: The proposed method for temporally resolved parametric assessment of Z-magnetization recovery (TOPAZ) is based on contiguous fast low-angle shot imaging readout after magnetization inversion from the pulsed steady state. Thereby, segmented k-space data are acquired over multiple heartbeats, before reaching steady state. This results in sampling of the inversion-recovery curve for each heart phase at multiple points separated by an R-R interval. Joint T1 and B1+ estimation is performed for reconstruction of cardiac phase-resolved T1 and B1+ maps. Sequence parameters are optimized using numerical simulations. Phantom and in vivo imaging are performed to compare the proposed sequence to a spin-echo reference and saturation pulse prepared heart rate–independent inversion-recovery (SAPPHIRE) T1 mapping sequence in terms of accuracy and precision. Results: In phantom, TOPAZ T1 values with integrated B1+ correction are in good agreement with spin-echo T1 values (normalized root mean square error = 4.2%) and consistent across the cardiac cycle (coefficient of variation = 1.4 ± 0.78%) and different heart rates (coefficient of variation = 1.2 ± 1.9%). In vivo imaging shows no significant difference in TOPAZ T1 times between the cardiac phases (analysis of variance: P = 0.14, coefficient of variation = 3.2 ± 0.8%), but underestimation compared with SAPPHIRE (T1 time ± precision: 1431 ± 56 ms versus 1569 ± 65 ms). In vivo precision is comparable to SAPPHIRE T1 mapping until middiastole (P > 0.07), but deteriorates in the later phases. Conclusions: The proposed sequence allows cardiac phase-resolved T1 mapping with integrated B1+ assessment at a temporal resolution of 40 ms. Magn Reson Med 79:2087–2100, 2018.
Bibliographical noteFunding Information:
Early techniques for the quantification of the longitudinal relaxation time (T1) in the heart have used continuous imaging using equidistant fast low-angle shot (FLASH) excitations, following an initial inversion pulse, as origi-nally proposed by Look and Locker (8). Although such approaches did not provide a pixel-wise T1 map due to the 1ElectricalandComputerEngineering,UniversityofMinnesota,Minneapolis, absence of electrocardiographic (ECG) triggering, they Minnesota,USA. allowed for a regional estimation of myocardial T1 using a 2Center for Magnetic Resonance Research, University of Minnesota, region of interest (ROI) analysis (9). For voxel-wise quantifi-3Minneapolis,Minnesota,USA. cation, the modified Look-Locker inversion-recovery HeidelbergUniversity,Mannheim,Germany.ComputerAssistedClinicalMedicine,UniversityMedicalCenterMannheim, sequence was introduced, performing single-shot imaging, 4Cardiovascular Division, Department of Medicine, University of Minnesota, triggered to the end-diastolic quiescence in a Look-Locker-Minneapolis, Minnesota, USA. type inversion-recovery experiment (10). This enabled spa-5Working Group on Cardiovascular Magnetic Resonance Imaging, Experi- tially resolved quantification of the T1 time as a parameter Delbru€ck-CentrumandCharite′-MedicalUniversityBerlin,Berlin,Germany.mentalandClinicalResearchCenter,Joint Cooperation oftheMax- map (T1 mapping) and established widespread use of quan-6DepartmentofCardiologyandNephrology,HELIOSKlinikumBerlin-Buch, titative cardiovascular MRI. Other imaging sequences, Berlin, Germany. based on inversion (11–13) or saturation (14,15) recovery or Grant Support: NIH, grant numbers R00HL111410 and P41EB015894; NSF, a combination of both (16,17), have been subsequently grantnumberCCF-1651825. introduced for myocardial T1 mapping, each offering a dis-SAPPHIREsequence.S.W.andM.A.areinventorsofanUSandEuropeanpatentdescribingthe tinct profile of advantages and disadvantages (18–20).
- + mapping
- T mapping
- cardiac imaging
- quantitative myocardial tissue characterization