Synopsis

Modified Look-Locker inversion recovery (MOLLI) uses bSSFP readout due to its high SNR, however, bSSFP is sensitive to off-resonance effects which result in banding artifacts. Recently, GRE has been proposed as an alternative readout for MOLLI, however, the low SNR efficiency of GRE-MOLLI is still a major problem. In this work, we propose to use a denoising reconstruction framework with low-rank and sparsity constraints to improve the low SNR of GRE-MOLLI. The proposed denoising method improved the low SNR of GRE-MOLLI, and multi-slice GRE-MOLLI is feasible for artifact-free T₁ mapping with wider spatial coverage at high magnetic fields ($$$\geq 3T$$$).

Introduction

Methods

Low-rank⁷ and sparsity constraints⁸ have been exploited for various applications including diffusion-weighted^9,10 and accelerated cardiac imaging¹¹. In this work, we have adapted the denoising reconstruction framework with low-rank and sparsity constraints to GRE-MOLLI case by formulating the problem as:

$$\hat{\rho}=arg\min_{\rho\in C^{N\times M}}\parallel d -\rho \parallel_2^2+\lambda\parallel\Psi(\rho)\parallel_1 s.t. rank(\rho)\leq L$$

where $$$d$$$ denotes measurement, $$$\rho$$$ denotes true noise-free GRE-MOLLI data, $$$\lambda$$$ denotes lagrangian multiplier, $$$\Psi$$$ denotes sparsifying transform, $$$C$$$ denotes the field of complex numbers, $$$N$$$ denotes number of voxels, and $$$M$$$ denotes number of time measurements. Low-rank constraint was used with the expectation that high spatio-temporal correlation exists between the noise-free, time-evolving voxel signals of GRE-MOLLI assuming homogenous T₁ within the myocardium and blood pool. Total variation¹² was used for to exploit the spatial correlation of myocardium and blood pool structure in the heart for effective edge-preserving smoothing. The above problem was solved similarly as in Ref.[11].

All experiments were performed on a PET-MR scanner (mMR Biograph, Siemens Healthcare, Erlangen, Germany) using body coil for transmission and body/spine matrix coils for signal reception. Simulations were performed using a steady-state spoiled GRE signal model¹³ to determine the Ernst angle of GRE readout for a range of T₁ values and compare the signal GRE with bSSFP prior to experiments. The performance of GRE-MOLLI sequence was tested using a T₁ phantom consisting of different concentrations of gadolinium (Gd) with simulated heart rate of 60beats/min prior to in vivo experiment. A porcine model was imaged under a study protocol approved by the IACUC. Cardiac images of an anesthetized porcine model were obtained at mid-ventricular short axis view with free-breathing using ECG-triggered single-shot acquisitions. Acquisitions were performed over 220ms window in end-diastole period of each cardiac cycle to minimize artifacts from cardiac motion.

2D multi-slice acquisitions were performed using MOLLI¹ with 5-(3)-5-(3) scheme (i.e. 5 acquisition per IR set and 3 additional cardiac cycles to enable full recovery of longitudinal magnetization) over 16 heartbeats for each slice, and inversion times (TI) of 100 and 200ms were used for each IR set. The general imaging parameters were: field-of-view (FOV) = 360$$$\times$$$304mm², matrix size = 192$$$\times$$$162, number of slices=16, and slice thickness=6 mm. GRAPPA¹⁴ and POCS¹⁵ reconstructions were performed prior to data analysis. T₁ was estimated for each voxel via least-square fitting to three-parameter IR signal model³ using variable projection algorithm¹⁶ and conventional LL equation for T₁ estimation¹⁷. Relative bias of T₁ from GRE-MOLLI was calculated using T₁ from bSSFP-MOLLI as reference³.

Results

Conclusion

Acknowledgements

References

1. Messroghli, Daniel R., et al. "Modified Look-Locker Inversion recovery (MOLLI) for high-resolution T1 mapping of the heart." Magnetic Resonance in Medicine 52.1 (2004): 141-146.

2. Scheffler, Klaus et al. "Principles and applications of balanced SSFP techniques." European Radiology 13.11 (2003): 2409-2418.

3. Cameron, Donnie et al. "Selection of magnetization catalyzation and readout methods for modified Look–Locker inversion recovery: AT 1 mapping primer." Magnetic Resonance Imaging 33.4 (2015): 363-373.

4. Shao, Jiaxin et al. "Myocardial T1 mapping at 3.0 tesla using an inversion recovery spoiled gradient echo readout and bloch equation simulation with slice profile correction (BLESSPC) T1 estimation algorithm." Journal of Magnetic Resonance Imaging 43.2 (2016): 414-425.

5. Rodgers, Christopher T. et al. "Inversion recovery at 7 T in the human myocardium: measurement of T1, inversion efficiency and B1+." Magnetic Resonance in Medicine 70.4 (2013): 1038-1046.

6. Jang, Jihye, et al. "Comparison of spoiled gradient echo and steady-state free-precession imaging for native myocardial T1 mapping using the slice-interleaved T1 mapping (STONE) sequence." NMR in Biomedicine 29.10 (2016): 1486-1496.

7. Liang, Zhi-Pei. "Spatiotemporal imaging with partially separable functions." 2007 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro. IEEE, 2007.

8. Haldar, Justin P., et al. "Spatiotemporal imaging with partially separable functions: A matrix recovery approach." 2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro. IEEE, 2010.

9. Haldar, Justin P., et al. "Improved diffusion imaging through SNR-enhancing joint reconstruction." Magnetic resonance in medicine 69.1 (2013): 277-289.

10. Lam, Fan, et al. "Denoising diffusion-weighted magnitude MR images using rank and edge constraints." Magnetic resonance in medicine 71.3 (2014): 1272-1284.

11. Zhao, Bo, et al. "Image reconstruction from highly undersampled-space data with joint partial separability and sparsity constraints." IEEE transactions on medical imaging 31.9 (2012): 1809-1820.

12. Chambolle, Antonin. "An algorithm for total variation minimization and applications." Journal of Mathematical imaging and vision 20.1-2 (2004): 89-97.

13. Haacke, E. Mark, et al. Magnetic resonance imaging: physical principles and sequence design. Vol. 82. New York:: Wiley-Liss, 1999.

14. Griswold, Mark A. et al. "Generalized autocalibrating partially parallel acquisitions (GRAPPA)." Magnetic Resonance in Medicine 47.6 (2002): 1202-1210.

15. Cuppen, Jan et al. "Reducing MR imaging time by one-sided reconstruction." Magnetic Resonance Imaging 5.6 (1987): 526-527.

16. Golub, Gene et al. "Separable nonlinear least squares: the variable projection method and its applications." Inverse Problems 19.2 (2003): R1.

17. Deichmann, R. et al. "Quantification of T 1 values by SNAPSHOT-FLASH NMR imaging." Journal of Magnetic Resonance 96.3 (1992): 608-612.

18. Lee, Jason J., et al. "Myocardial T1 and extracellular volume fraction mapping at 3 tesla." Journal of cardiovascular magnetic resonance 13.1 (2011): 1.