Native and Post-Contrast Mapping, Cine Imaging, and LGE Imaging Using Cine MR Fingerprinting: Initial Experience for Cardiomyopathy Phenotyping (RF_SA_495)

Background: Cine MR Fingerprinting (MRF) has the potential for ‘all-in-one’ assessment of myocardial function and tissue characteristics by providing cardiac phase-resolved maps of T₁, T₂, M₀, and B_1⁺.^1,2 To date, cine MRF has primarily been applied to native mapping in healthy subjects; however, post-contrast mapping provides additional information and enables synthetic LGE imaging, as demonstrated by previously using diastolic (single-phase) MOLLI,³ MRF,⁴ and SASHA.⁵ The goal of this study was: (1) to use post-contrast cine MRF to obtain functional (cine) LGE images, (2) to evaluate the feasibility of using cine MRF as a single-sequence platform for native and post-contrast mapping, cine imaging, and LGE imaging in cardiomyopathy patients.

Methods: Three healthy volunteers and five patients with ischemic, non-ischemic, and infiltrative cardiomyopathy were imaged at 1.5T (MAGNETOM Aera, Siemens). Conventional bSSFP cine (short-axis stack), tissue property maps (3 slices) using MOLLI⁶ and T₂-prepared bSSFP,⁷ and PSIR LGE images (short-axis stack) were acquired. Additionally, a whole-heart short-axis stack of cine MRF data was acquired both pre-contrast and 10 min after gadolinium contrast administration. Cine MRF data were collected using a 2D breathheld and retrospectively ECG-gated (25 phases) FISP sequence.¹ Each slice was acquired during a 10.7s breathhold, with multiple breathholds performed to achieve whole-LV coverage with 9-12 slices at 1.6x1.6x8.0 mm³. T₁, T₂, and M₀ maps were obtained using a Deep Image Prior reconstruction with B₁⁺ spin history correction.⁸ Synthetic bright- and dark-blood cine images were calculated to evaluate cardiac volumes and function. Additionally, post-contrast cine MRF maps were used to generate synthetic PSIR LGE images. LVEDV, LVESV, LVEF, and native T₁ and T₂ in the LV septum were compared with values from conventional sequences using Bland-Altman analyses.

Results: The time needed to acquire 12 short-axis cine MRF slices was ~4 min. Maps, cine images, and LGE images from cine MRF scans are shown in Fig. 1 for a healthy subject and patients with ischemic and nonischemic cardiomyopathy. Volumetric analyses showed good agreement with standard scans (Fig. 2), with biases of -29.8 mL, -19.1 mL, and 1.6% for LVEDV, LVESV, and LVEF, respectively. There was a trend towards higher native T₁ (bias 41 ms), and significantly lower native T₂ (bias -6.1 ms) using cine MRF, consistent with prior studies.⁸ Fig. 3 shows cardiac phase-resolved LGE images derived from post-contrast cine MRF maps for patients with and without abnormal enhancement.

Conclusion: Cine MRF demonstrated good agreement with standard sequences for assessing myocardial volumes, function, and LGE. Additionally, this work showed the feasibility of using MRF to generate functional LGE images, which could enable assessment of scar motility, as in related work by Weingartner et al.⁹ Future work will continue to evaluate cine MRF as an ‘all-in-one’ platform in a larger cohort of cardiomyopathy patients.