Initial feasibility of free-breathing co-registered cardiac T1, T2 and ADC mapping using echo planar imaging with whole left ventricle coverage (RF_TH_278)

Quantitative cardiac MRI offers non-invasive characterization of myocardial tissue composition for the diagnosis of cardiomyopathies. The use of separate breath-holding 2D scans for myocardial T1 and T2 is the current clinical practice. However, there is a clinical need¹ for whole-heart coverage to characterize complex regional distribution patterns of disease. We previously proposed MP-EPI, a free-breathing method to produce co-registered cardiac T1, T2, and ADC mapping using single shot EPI in 3 slices requiring 10 minutes. Slice interleaved acquisition has demonstrated accurate and precise quantification of native myocardial T1 maps with the additional benefit of efficiently covering the entire LV². In this study, we propose a multi-slice T1, T2, and ADC mapping sequence design with interleaved acquisition of slices for complete LV coverage using non-selective inversion preparation, varying TE times, and motion compensated (M2) diffusion gradients.

Methods:

Imaging was performed on a 3T MRI scanner (MAGNETOM Cima.X, Siemens Healthineers AG, Forchheim, Germany) in 5 healthy volunteers.

The proposed method using an M2³, Resolution: 2.7x2.7x8 mm³, BW=3005 Hz, EPI Factor: 48, Acquisition windows: 20 ms, 8 mm slice thickness, 2 mm gap) achieves free breathing co-registered T1, T2, and ADC mapping of 12 LV slices by performing 43 repeated acquisitions (RAs) in an interleaved slice order (Figure 1). Deformable motion correction is used to correct for in-plane motion⁴. T1, T2, and ADC maps are fitted separately using mono-exponential curve fitting. For comparison, 3-slice T1-MOLLI and T2-FLASH maps are acquired in systole and acquisition windows are reduced to minimize cardiac motion⁵ (Resolution: 2.0x2.0x8 mm³, FOV: 300x300 mm², Acquisition windows: 150 ms).

Results:

Average acquisition time was 8.9 ± 1.26 minutes. Figure 2A shows raw images in the basal slice. Figure 2B shows whole heart T1/T2/ADC maps from the same healthy subject. Global T1 values in the myocardium (Figure 3B) were not significantly different (p=0.11) between the proposed method: 1321.6±52.7 ms and MOLLI: 1256.9±46.5 ms with the limit of agreement of ±266 ms. T2 values were significantly lower (p=0.021) between the proposed method: 39.4±1.4 ms and T2prep FLASH: 41.4±2.0 ms with the limit of agreement of ± 4.9 ms. Global ADC values were within normal range (1.50±0.04 µm²/ms) from literature⁶.

Conclusion:

We demonstrated initial feasibility of a free breathing co-registered T1/T2/ADC mapping method in less than 10 minutes using echo planar imaging with whole LV coverage in a small group of healthy volunteers. This method has the potential to improve clinical analysis for cardiomyopathy involving focal heterogeneity by providing regional characteristics of the myocardial tissue.