Comparison Between Spin Echo and Stimulated Echo Diffusion Encoding for Diffusion-Weighted Cardiac Magnetic Resonance (DW-CMR) at 3T
Christopher Nguyen1, Peter Speier2, Xiaoming Bi3, and Debiao Li1,4

1Biomedical Imaging Research Institute, Cedars-Sinai Medical Center, Los Angeles, CA, United States, 2Siemens Healthcare GmbH, Erlangen, Germany, 3Siemens Healthcare, Los Angeles, CA, United States, 4Bioengineering, University of California Los Angeles, Los Angeles, CA, United States

Synopsis

Clinical application of diffusion-weighted cardiac magnetic resonance (DW-CMR) has demonstrated promise in detecting and characterizing the tissue microphysiology and/or microstructure. The aim of this study is to compare a recently developed SE diffusion-prepared approach with a conventional STE DW-CMR technique in normal volunteers on a 3T system. M2-SE derived MD values were significantly higher than STE derived values despite the same prescribed b-value and the difference is not dependent on heart rate. Both DW-CMR techniques were reproducible and motion corruption was seen to be overall less in M2-SE.

Introduction

Clinical application of diffusion-weighted cardiac magnetic resonance (DW-CMR) has demonstrated promise in detecting and characterizing the tissue microphysiology and/or microstructure of myocarditis, myocardial edema, and myocardial infarction1-3. However, as new DW-CMR techniques4 emerge beyond the conventional DW-CMR5, there is a clear difference in the absolute quantification of mean diffusivities (MD) dependent on whether the DW data was collected with spin echo (SE) or stimulated echo (STE) encoding. The aim of this study is to compare a recently developed SE diffusion-prepared approach with a conventional STE DW-CMR technique in normal volunteers on a widely available clinical 3T system (MAGNETOM Skyra, Siemens Healthcare, Germany).

Methods

A recently developed second order motion compensated (M2) SE diffusion-prepared approach (10 avg, TRg=2RR, TEprep = 67ms, free-breathing, 2D single shot bSSFP readout: TR/TE = 2.7/1.4ms, FA = 90°, R = 2, partial Fourier factor = 6/8) was compared with a conventional STE DW-CMR technique (10 avg, breath-hold, 2D SS EPI readout: TR/TE = 2RR/20ms, reduced FOV factor = 5, partial Fourier factor = 6/8) in healthy volunteers (N=5). Prototype sequences implementing M2-SE and STE DW-CMR were both applied in mid diastole to avoid strain contamination and reduce bulk motion during encoding. Both DW-CMR techniques acquired 3 non-collinear diffusion directions at a b-value = 350 s/mm2 and a single b-value = 30 s/mm2 image at the mid ventricular short axis slice. Total scan time for both techniques were kept at ~5 minutes. Regional (6 AHA segments) myocardial MD values (N=30 segments) were tested for significant differences using paired t-test. Intra-scan reproducibility was tested interleaving M2-SE with STE measurements and alternating the initial measurement to reduce bias. Bland-Altman plots were generated to qualitatively test for Type 1 and 2 errors. Motion corruption defined by > 10% myocardium exhibiting signal void within each AHA segment for each DW measurement (N=90) was tallied and an overall percentage was reported. Data deemed motion corrupt were removed from MD estimation. Dependency on heart rate was also examined.

Results

Heart rate for all healthy volunteers were stable (72±7 BPM) throughout the 5 minutes of scanning for each measurement. MD values were significantly (p=0.02) higher in M2-SE (1.5±0.3 um2/ms) compared with STE (1.2±0.2 um2/ms). Regional analysis of MD values showed lateral wall variability to be significantly (p=0.03) higher in M2-SE than STE (SD: 0.5 vs 0.2 um2/ms, respectively). The mean bias in MD between M2-SE and STE was relatively constant (0.4±0.2 um2/ms) and had weak correlation with heart rate (r2=0.1). Both methods demonstrated high intra-scan reproducibility with low bias in repeated measures of MD (M2-SE: 0.2 um2/ms; STE: 0.1 um2/ms). Overall, M2-SE exhibited significantly less motion corruption than STE (7.8% vs 11%, respectively). However, STE demonstrated significantly less motion corruption than M2-SE (22% vs 13%, respectively) at higher heart rates (>80 BPM).

Discussion

The disparity between MD estimations between M2-SE and STE is consistent with other groups’ findings for brain, skeletal muscle, and heart 6-8. These studies hypothesize that the large difference in diffusion mixing times may be the source for the disparity. Our work is supportive of this claim demonstrating the disparity in MD estimation is not dependent on heart rate, which may have been confounding source for in vivo DW-CMR imaging. Another confounding source may be STE’s dependency on strain, which in this study was mitigated by “sweet spot” triggering. A more complete future comparison study would include multiple triggering throughout the cardiac cycle.

Conclusion

M2-SE derived MD values were significantly higher than STE derived values despite the same prescribed b-value and the difference is not dependent on heart rate. Both DW-CMR techniques were reproducible and motion corruption was seen to be overall less in M2-SE. Higher stable heart rates decreased M2-SE’s motion robustness and caused more variations in lateral wall MD values compared with STE.

Acknowledgements

1F31EB018152-01A1

References

1. Pop, M. et al. Quantification of fibrosis in infarcted swine hearts by ex vivolate gadolinium-enhancement and diffusion-weighted MRI methods. Phys. Med. Biol. 58, 5009–5028 (2013).

2. Nguyen, C. et al. In vivo contrast free chronic myocardial infarction characterization using diffusion-weighted cardiovascular magnetic resonance. Journal of Cardiovascular Magnetic Resonance 16, 1770 (2014).

3. Potet, J. et al. Detection of myocardial edema with low-b-value diffusion-weighted echo-planar imaging sequence in patients with acute myocarditis. Radiology 269, 362–369 (2013).

4. Nguyen, C. et al. In vivo three-dimensional high resolution cardiac diffusion-weighted MRI: a motion compensated diffusion-prepared balanced steady-state free precession approach. Magn. Reson. Med. 72, 1257–1267 (2014).

5. Nielles-Vallespin, S. et al. In vivo diffusion tensor MRI of the human heart: Reproducibility of breath-hold and navigator-based approaches. Magn. Reson. Med. 70, 454–465 (2012).

6. Noehren, B., Andersen, A., Feiweier, T., Damon, B. & Hardy, P. Comparison of twice refocused spin echo versus stimulated echo diffusion tensor imaging for tracking muscle fibers. J. Magn. Reson. Imaging 41, 624–632 (2014).

7. Heiland, S., Dietrich, O. & Sartor, K. Diffusion-weighted imaging of the brain: comparison of stimulated- and spin-echo echo-planar sequences. Neuroradiology 43, 442–447 (2001).

8. Deuster, von, C., Stoeck, C. T., Genet, M., Atkinson, D. & Kozerke, S. Spin echo versus stimulated echo diffusion tensor imaging of the in vivo human heart. Magn. Reson. Med. n/a–n/a (2015). doi:10.1002/mrm.25998

Figures

Figure 1 - Point plot demonstrating the difference between M2-SE and STE in deriving MD. M2-SE-derived MD values were significantly increased than STE-derived values.

Figure 2 - Representative raw images from a single average for STE acquired DW-CMR images. The last column represents an image that was scored as motion corrupted in both the anteroseptal and inferoseptal AHA segments.

Figure 2 - Representative raw images from a single average for M2-SE acquired DW-CMR images. The last column represents an image that was scored as motion corrupted in inferolateral AHA segment.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3110