Introduction

High-performance 0.55T MRI provides the opportunity to evaluate lung anatomy and function with excellent image quality, largely due to reduced magnetic susceptibility^1,2, longer T₂ and T₂^*3. The longer T₂^* opens up exciting new opportunities including long-readout (5-10msec) stack-of-spiral imaging^4,5. However, only a limited number of studies³ has explored lung parameter mapping at low field and information about lung MR parameters can potentially provide guidance for advancing and optimizing the pulse sequence design. $$$R_{2}$$$ is typically estimated from spin-echo pulse sequences or T2-prepared bSSFP pulse sequences while $$$R_{2}^{\ast}$$$ is typically estimated from multi-echo gradient-echo pulse sequences^6,7. Due to the intrinsic difference between pulse sequences^8,9 and different acquisition parameters in separate $$$R_{2}$$$ and $$$R_{2}^{\ast}$$$ mapping, various contributions can lead to bias in the $$$R_{2}^{\prime}$$$ estimate which is more informative in some applications^10,11.
At low field, the application of turbo spin echo (TSE) has provided excellent imaging quality and diagnostic value¹². In this work, we utilize one pulse sequence, a custom echo-shifted TSE, to jointly estimate off-resonance $$$\Delta f$$$, $$$R_{2}$$$, and $$$R_{2}^{\prime}$$$, which are transverse relaxation rate due to thermodynamics (irreversible) and field inhomogeneity (reversible)¹³.

METHODS

Data acquisition: Experiments were performed using a whole body 0.55T system (prototype MAGNETOM Aera, Siemens Healthineers, Erlangen, Germany) equipped with high-performance shielded gradients (45 mT/m amplitude, 200 T/m/s slew rate). A custom ECG triggered, echo-shifted TSE pulse sequence was performed on 4 healthy volunteers (2M/2F, Age 24-28). Imaging parameters: spin-echo times (t_SE): 26, 39, and 52 ms; echo shift times (t_shift): -1.7, -1, 0, 1, and 1.4 ms for t_SE = 26 ms and 39 ms; echo spacing = 13 ms; FOV = 360 x 270 mm²; matrix size = 128 x 96; slice thickness = 6 mm; bandwidth = 260 Hz/pixel; echo train length = 15; ECG trigger delay = 550 ms for imaging at diastole. Each breath hold was about 7 to 8 seconds depending on the heartrate. Total scan time was about 5 minutes per subject. Figure 1 shows the pulse sequence diagram animated to include t_shift = -1.7, -1, 0, 1 and 1.4 ms.
Joint estimation of $$$R_{2}$$$, $$$R_{2}^{\prime}$$$, $$$\Delta f$$$ : A 2D Gaussian filter (σ = 0.7 pixel size = 1.33 mm, kernel size = 5) was applied on complex-valued images for smoothing before parameter estimation. Spatial registration was not applied. We used a three-parameter complex signal model based on 180° spin echoes for $$$R_{2}$$$ and $$$R_{2}^{\prime}$$$ estimation. Two equations describing the signal evolution for rephasing ($$$t_{shift }\leq0$$$) and dephasing ($$$t_{shift }\geq0$$$) portions of spin echoes were used to jointly estimate $$$R_{2}$$$, $$$R_{2}^{\prime}$$$, and $$$\Delta f$$$ :
\begin{array}{} \mathrm{S}\left(\vec{r}, t_{shift}\right)=\mathrm{S}_{0} e^{-R_{2}(\vec{r})\left(t_{SE}+t_{shift}\right)+R_{2}^{\prime}(\vec{r}) t_{shift}} e^{-j 2 \pi \Delta f(\vec{r}) t_{shift }}&{for\ t_{shift} \leq0} \\ {\mathrm{~S}\left(\vec{r},t_{shift}\right)=\mathrm{S}_{0} e^{-R_{2}(\vec{r})\left(t_{SE}+t_{shift}\right)-R_{2}^{\prime}(\vec{r}) t_{shift}} e^{-j 2 \pi \Delta f(\vec{r}) t_{shift}}} &{for\ t_{shift} \geq0}\end{array}
where $$$\mathrm{S}\left(\vec{r}, t_{shift}\right) \in \mathbb{C}$$$ denotes a complex-valued signal at $$$\vec{r}$$$ position acquired at t_shift, $$$S_{0} \in \mathbb{C}$$$ is complex-valued spin magnetization, and $$$R_{2} \in \mathbb{R}$$$ and $$$R_{2}^{\prime} \in \mathbb{R}$$$ are real-valued transverse relaxation rates and is static off-resonance. Three spin-echoes and four echo shift times were used. Parameter estimation was performed using a nonlinear least square approach with non-negative real constraints on $$$R_{2}$$$ and $$$R_{2}^{\prime}$$$, and real constraint on $$$\Delta f$$$ .

RESULTS

Table 1 summarizes the results from four healthy volunteers. The estimated R₂ of lung parenchyma was approximately 23.4 s^­-1 ($$$T_{2}$$$ = 42.7 ms) and was slightly lower than that reported at 1.5T¹². The estimated $$$R_{2}^{\prime}$$$ of lung parenchyma was approximately 53.9 s^­-1 ($$$T_{2}^{\prime}$$$= 18.6 ms) and the resulting estimated $$$R_{2}^{\ast}$$$ was approximately 77.6 Hz ($$$T_{2}^{\ast}$$$ =12.9ms), which is 10 times lower than that reported at 1.5T^3,14.
Figures 3 and 4 show $$$R_{2}$$$, $$$R_{2}^{\prime}$$$ and $$$\Delta f$$$ maps and histograms from four volunteers. The off resonance of lung parenchyma ranged from 30 Hz to 80 Hz (3.41ppm) and the mean $$$\Delta f$$$ is 46.5 Hz (1.95ppm).

DISCUSSION

A previous study³ utilized two different sequences with different contrasts (T2-prepared bSSFP and multi-echo GRE) to estimate $$$R_{2}$$$ and $$$R_{2}^{\ast}$$$, which we suspected could lead to errors when estimating $$$R_{2}^{\prime}$$$. Our single-sequence approach produced comparable $$$R_{2}$$$ and $$$R_{2}^{\ast}$$$ values, dispelling this concern. The reduced $$$R_{2}$$$ and $$$R_{2}^{\ast}$$$ estimates at low field allow substantial new flexibility in pulse sequence design, including the use of long-readout spiral imaging, multi-echo imaging, and balanced steady-state free precession.
This study has several limitations. First, the total readout time (i.e., 200 ms) used in this study results in nonnegligible T2 blurring, for $$$R_{2}$$$ around 23.4 s^-1. T2 blurring correction can potentially improve the accuracy of quantification^15,16. Second, the assumption of perfect B1⁺ homogeneity in the signal model may lead to bias in quantification. Advanced signal modeling along with a separate B1⁺ map could result in improved quantification¹⁷.

CONCLUSION

With a custom ECG-triggered echo-shifted TSE pulse sequence, simultaneous estimation of $$$R_{2}$$$, $$$R_{2}^{\prime}$$$, and $$$\Delta f$$$ was feasible at 0.55T. The mean of $$$R_{2}$$$ is ~23.4 s^-1 and $$$R_{2}^{\prime}$$$ is ~53.9 s^-1(18.6 ms). $$$R_{2}$$$ is slightly lower as expected, while $$$R_{2}^{\prime}$$$ and $$$R_{2}^{\ast}$$$ are significantly lower at 0.55T compared with 1.5T. Off-resonance is up to 3.41ppm which is also consistent with previous results at 1.5T¹⁸. The quantitative parameters found in this work could benefit sequence design for lung MRI at low field.

Acknowledgements

We acknowledge grant support from the National Science Foundation (#1828736) and research support from Siemens Healthineers.

References

1. Bhattacharya I, Ramasawmy R, Javed A, et al. Oxygen-enhanced functional lung imaging using a contemporary 0.55 T MRI system. NMR Biomed. 2021;34(8):1-10. doi:10.1002/nbm.4562
2. Campbell-Washburn AE, Bhattacharya I, Moss J, et al. T2-Weighted Lung Parenchymal Imaging Using a High-Performance Low Field MRI System. 2020:A3250-A3250. doi:10.1164/ajrccm-conference.2020.201.1_meetingabstracts.a3250
3. Campbell-Washburn AE, Ramasawmy R, Restivo MC, et al. Opportunities in Interventional and Diagnostic Imaging by Using High-Performance Low-Field-Strength MRI. Radiology. 2019;293(2):384-393. doi:10.1148/radiol.2019190452
4. Javed A, Ramasawmy R, Moss J, et al. Optimized 3D spiral ultra-short echo time free-breathing pulmonary imaging on a high-performance low-field 0.55T scanner. Proc Intl Soc Mag Reson Med 29. 2021:0596.
5. Javed A, Ramasawmy R, Su P, Benkert T, Majeed W, Campbell-Washburn AE. Free-breathing oxygen- enhanced pulmonary imaging with self-gated stack- of-spirals ultra-short echo time sequence at 0.55T. Ismrm 2021. 2021:3238.
6. Hatabu H, Alsop DC, Listerud J, Bonnet M, Gefter WB. T2* and proton density measurement of normal human lung parenchyma using submillisecond echo time gradient echo magnetic resonance imaging. Eur J Radiol. 1999;29(3):245-252. doi:10.1016/S0720-048X(98)00169-7
7. Pracht ED, Arnold JFT, Wang T, Jakob PM. Oxygen-enhanced proton imaging of the human lung using T2*. Magn Reson Med. 2005;53(5):1193-1196. doi:10.1002/mrm.20448
8. Chavhan GB, Babyn PS, Jankharia BG, Cheng HLM, Shroff MM. Steady-state MR imaging sequences: Physics, classification, and clinical applications. Radiographics. 2008;28(4):1147-1160. doi:10.1148/rg.284075031
9. Kacere RD, Pereyra M, Nemeth MA, Muthupillai R, Flamm SD. Quantitative assessment of left ventricular function: Steady-state free precession MR imaging with or without sensitivity encoding. Radiology. 2005;235(3):1031-1035. doi:10.1148/radiol.2353030995
10. Wood JC, Enriquez C, Ghugre N, et al. MRI R2 and R2* mapping accurately estimates hepatic iron concentration in transfusion-dependent thalassemia and sickle cell disease patients. Blood. 2005;106(4):1460-1465. doi:10.1182/blood-2004-10-3982
11. Siemonsen S, Fitting T, Thomalla G, et al. T2′ imaging predicts infarct growth beyond the acute diffusion-weighted imaging lesion in acute stroke. Radiology. 2008;248(3):979-986. doi:10.1148/radiol.2483071602
12. Bhattacharya I, Ramasawmy R, Javed A, et al. Assessment of Lung Structure and Regional Function Using 0.55 T MRI in Patients With Lymphangioleiomyomatosis. Invest Radiol. 2021;Publish Ah(00):1-9. doi:10.1097/rli.0000000000000832
13. Ma J, Wehrli FW. Irreversible Contribution to the Transverse-Relaxation Rate. Integr Vlsi J. 1996;69:61-69.
14. Wild JM, Marshall H, Bock M, et al. MRI of the lung (1/3): Methods. Insights Imaging. 2012;3(4):345-353. doi:10.1007/s13244-012-0176-x
15. Zhou X, Liang Z ‐P, Cofer GP, Beaulieu CF, Suddarth SA, Johnson GA. Reduction of ringing and blurring artifacts in fast spin‐echo imaging. J Magn Reson Imaging. 1993;3(5):803-807. doi:10.1002/jmri.1880030518
16. Tamir JI, Uecker M, Chen W, et al. T2 shuffling: Sharp, multicontrast, volumetric fast spin-echo imaging. Magn Reson Med. 2017;77(1):180-195. doi:10.1002/mrm.26102
17. Chhetri G, McPhee KC, Wilman AH. Bloch modelling enables robust T2 mapping using retrospective proton density and T2-weighted images from different vendors and sites. Neuroimage. 2021;237(May):118116. doi:10.1016/j.neuroimage.2021.118116
18. Bergin CJ, Glover GH, Pauly JM. Lung parenchyma: Magnetic susceptibility in MR imaging. Radiology. 1991;180(3):845-848. doi:10.1148/radiology.180.3.1871305