Pre- and Post-contrast Joint T1/T2 Mapping at 0.55T for Detection of Myocardial Infarction: Validation in Porcine Models (RF_TH_189)

Background: CMR T1 and T2 mapping has been widely used in clinic for the detection of various cardiac diseases [1]. Recently, there is increased interest in low-field-strength MRI systems, which show promise to benefit CMR imaging with the advantages of lower cost, lower SAR and fewer field inhomogeneities [2-4]. However, most of the parametric mapping sequences have been proposed and validated at common field strengths such as 1.5T and 3T. We recently developed a free-breathing 3D isotropic resolution joint T1/T2 mapping research sequence at 0.55T, which improved the efficiency of whole-heart imaging compared with conventional 2D sequences [5]. In this study, we sought to validate the feasibility of 3D joint T1/T2 for imaging of myocardial infarction (MI) in porcine models before and after contrast injection. 

Methods: All MR experiments were performed on a 0.55T MR scanner (MAGNETOM Free.Max; Siemens Healthineers AG, Erlangen, Germany). All animal studies were approved by the local Institutional Animal Care and Use Committee.
Three swine (approximately 2-3 months old at baseline and weighing 30-40 kg over the study duration) were included in this study. CMR imaging was performed at three time points including baseline before MI, and at 1-week and 5-week post MI. 2D sequences including MOLLI, T2-prep bSSFP and phase sensitive inversion recovery (PSIR) late gadolinium enhancement (LGE) were used to acquire pre-contrast T1 and T2 maps, post-contrast reference T1 maps and LGE images [4]. A total dose of 0.2 mmol/kg of gadobutrol (Gadavist, Bayer Healthcare, Whippany, NJ) was used. 3D joint T1/T2 mapping was also performed both pre- and post-contrast with the same imaging parameters as in previous study [5]. 3D synthetic LGE images were generated from post-contrast 3D joint T1/T2 maps using Bloch equation-based signal simulation [6].
ROI of remote myocardium and MI was manually defined in images reformatted to mid short axis views to analyse the related T1 and T2 values. The left ventricle was segmented to generate bull’s eye plots of the myocardium.

Results: Representative T1 and T2 maps and LGE images from one swine at the three time points indicate that 3D joint T1/T2 mapping sequence provides comparable image quality and consistent MI detection as 2D images (Fig. 1). Fig. 2 shows bull’s eye plots of the same swine shown in Fig. 1 at 5-week and demonstrates that 3D joint T1/T2 mapping enables high-resolution visualization of the MI area. Quantitative analysis of T1/T2 values of remote myocardium and MI for all animals are shown in Fig. 3. At both 1-week and 5-week time points, 2D and 3D synthetic LGE images both show enhancement in the area of MI, where pre-contrast T2 increased and post-contrast T1 decreased. Pre-contrast T1 of MI is less altered at 1-week, but it increased at the 5-week time point. 2D and 3D joint T1/T2 maps showed similar trends for all the T1/T2 values.

Conclusion: The preliminary results indicate that the proposed 3D joint T1/T2 mapping sequence can efficiently acquire pre- and post-contrast T1 maps, T2 maps and generate synthetic LGE images at 0.55T that are in good agreement with corresponding 2D sequences. 3D imaging with high isotropic resolution supports multi-planar reformatting and may thereby improve the depiction of MI area over the whole heart. Validations in a larger cohort will be included in the future.