0238

Quantitative Sodium MRI of the Human Kidney at 7T Applying Respiratory Sorting

Anna K. Scheipers¹, Johanna Lott¹, Armin M. Nagel^1,2, Peter Bachert¹, Mark E. Ladd¹, and Tanja Platt¹
¹Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, ²University Hospital Erlangen, Institute of Radiology, Friedrich‐Alexander‐Universität Erlangen‐Nürnberg (FAU), Erlangen, Germany

Synopsis

Sodium (²³Na) plays an important role in many cellular processes, making it an interesting nuclei to investigate using MRI. Low in-vivo signals and short relaxation times require high magnetic field strengths, dedicated hardware and pulse sequences as well as correction methods to obtain reliable tissue sodium concentrations. Therefore we compare the influence of different correction methods on ²³Na concentration investigations in the human kidney and validate our methods with phantom measurements. We employ retrospective respiratory self‐gating for the quantitative ²³Na images as well as for the acquired B₁⁺ maps to reduce the influence of image blurring due to motion artifacts.

Introduction

The kidneys play a vital role in maintaining and balancing the fluid and electrolyte homeostasis in the human body, for which the extracellular sodium (²³Na) control in the kidneys is essential. The challenges of sodium MRI of the torso are not only those of sodium imaging in general such as relatively low in-vivo concentrations, very short transversal relaxation times and the low NMR sensitivity of the ²³Na nucleus compared with hydrogen (¹H), resulting in a relatively low signal-to-noise ratio (SNR) in ²³Na MR images, but also blurring due to respiratory and cardiac motion. Nevertheless, sodium MRI of the kidney has the potential to become a non-invasive biomarker for renal tissue function and viability¹. To achieve this, a reliable way for data acquisition as well as data post processing is needed to account for factors influencing the determined ²³Na concentration. For this we compare the influence of different correction methods on ²³Na concentration investigations in the human kidney and validate our methods with phantom measurements. We employ retrospective respiratory self‐gating for the quantitative ²³Na images as well as for the acquired B₁⁺ maps to reduce the influence of image blurring due to motion artifacts.

Methods

Quantitative sodium abdominal images were acquired for a phantom filled with NaCl solution (35mM) and one healthy volunteers (female, age: 31years, weight: 80kg, measured head first supine with hands above the head) on a 7T whole-body MR system (MAGNETOM 7T, Siemens Healthcare GmbH, Erlangen, Germany) with an oval-shaped birdcage coil with a large field of view of (400mm)³, custom-made for torso examinations². Density-adapted 3D radial sampling³ and a golden angle projection distribution⁴ were applied with the following parameters: flip angle α=61°, TE/TR=1ms/150ms, nominal spatial resolution (∆x)³=(5mm)³, pulse duration t_P=1.8ms, readout time t_RO=5ms, number of projections n_Proj=16300. For concentration determination four reference vials with NaCl solutions (20mM, 40mM, 60mM, 120mM) were positioned under the back of volunteer to minimize the influence of respiratory motion and to ensure a consistent placing across all measurements. For the phantom measurements the vial concentrations were also chosen to fit the range of investigated concentrations (20mM, 30mM, 40mM, 50mM). To generate B₁⁺ maps, the dual-angle method⁵ was applied (α=45°/90°, TE/TR=1.55ms/150ms, (∆x)³=(10mm)³ , t_P=3ms, t_RO=10ms, n_Proj=8200). The resulting total acquisition time was about 80min.
Since the used coil does not share the exact same hardware for transmitting and receiving, the principle of reciprocity cannot be applied^2,6. Thus, B₁^- maps of the applied radiofrequency (RF) coil were simulated in a phantom model (ε_r'=54, σ=0.43Sm^-1) and in a human voxel model^7,8 using the electromagnetic field simulation software CST Studio Suite 2019 (Dassault Systèmes, Vélizy-Villacoublay, France).
For both, spin density‐weighted images and B₁⁺ maps, retrospective respiratory self‐gating was applied, sorting the projections into two respiratory states (inhaled and exhaled) based on the respiration-dependent signal changes in the k-space center².
Datasets for B₁⁺ maps were Gauss-filtered (σ=10mm), image data was filtered with a Hamming filter.
Full datasets and respiratory sorted datasets were reconstructed and B₁⁺ corrected separately (see Fig.2) and the concentration was determined with a semi-automated segmentation tool developed in Matlab (The MathWorks Inc., Natick, MA, USA) using the known concentration in the reference vials and a linear fit. For the in-vivo data, region of interest (roi) for the whole kidneys were segmented in Matlab as well.

Results

In-vivo measurements in a healthy volunteer yielded a sodium concentration for the different motion states (all projections, inhaled, exhaled) between (72.6±20.8)mmol/L and (78.3±18.7)mmol/L for the left and (98.7±24.5)mmol/L and (101.8±30.7)mmol/L for the right kidney as shown in Tab.2.
The B₁⁺ and B₁^- corrections lead to a decrease in the standard deviation in both phantom and in-vivo measurements. In case of the phantom data, the corrections bring the determined concentration closer to the ground truth (see Tab.1). Motion sorted B₁⁺ correction was shown to be feasible with the applied setup for ²³Na MRI at 7T.

Discussion & Conclusion

Quantitative ²³Na MRI of the kidney benefits from high magnetic field strengths B₀⁹, granting a higher SNR. The simulated receive-field map for the phantom was not considering the exact setup. This may lead to the inhomogeneous concentration map in the homogeneous phantom and an overestimated concluded concentration when only evaluating parts of the phantom as performed using the segmented rois from the in-vivo measurements in Tab.2. More precise simulation data could homogenize the phantom concentration map and improve this result.
The human kidney contains different tissue types including cortex, medulla and renal pelvis¹⁰, therefore the here determined concentrations are average values for the combination of these tissue types. Registration of the ²³Na images with high-resolution hydrogen images would help to segment the different tissue compartments, therefore enabling a more accurate determination of the sodium concentration of each individual tissue type. Nevertheless the mesasured ²³Na concentrations are in good agreement with the literature measurements conducted at 3T¹¹, ranging from (48.6±5.3)mmol/L to (108.0±10.9)mmol/L, depending on the physiologic condition and segmented tissue type.
Moreover, motion compensation could be beneficial in ²³Na abdominal image reconstruction to reduce blurring and could be implemented in the future for more accurate quantitative measurements.

Acknowledgements

No acknowledgement found.

References

1. Zöllner, F. G., Konstandin, S., Lommen, J., Budjan, J., Schoenberg, S. O., Schad, L. R., and Haneder, S. (2016) Quantitative sodium MRI of kidney. NMR Biomed., 29: 197– 205., https://doi.org/10.1002/nbm.3274.

2. Platt, T, Umathum, R, Fiedler, TM, et al. In vivo self‐gated 23Na MRI at 7 T using an oval‐shaped body resonator. Magn Reson Med. (2018); 80: 1005– 1019. https://doi.org/10.1002/mrm.27103

3. Nagel, A.M., Laun, F.B., Weber, M.‐A., Matthies, C., Semmler, W. and Schad, L.R. (2009), Sodium MRI using a density‐adapted 3D radial acquisition technique. Magn. Reson. Med., 62: 1565-1573. https://doi.org/10.1002/mrm.22157

4. Chan, R.W., Ramsay, E.A., Cunningham, C.H. and Plewes, D.B. (2009), Temporal stability of adaptive 3D radial MRI using multidimensional golden means. Magn. Reson. Med., 61: 354-363. https://doi.org/10.1002/mrm.21837

5. E.K. Insko, L. Bolinger, Mapping of the Radiofrequency Field, Journal of Magnetic Resonance, Series A, Volume 103, Issue 1, 1993,Pages 82-85, ISSN 1064-1858, https://doi.org/10.1006/jmra.1993.1133.

6. Lott, J, Platt, T, Niesporek, SC, et al. Corrections of myocardial tissue sodium concentration measurements in human cardiac 23Na MRI at 7 Tesla. Magn Reson Med. 2019; 82: 159– 173. https://doi.org/10.1002/mrm.27703

7. Christ A, Kainz W, Hahn EG, et al. The Virtual Family—development of surface-based anatomical models of two adults and two children for dosimetric simulations. Phys Med Biol. 2010;55:N23-N38.

8. Hasgall PA, Di Gennaro F, Baumgartner C, Neufeld E, Gosselin MC, Payne D, Klingenböck A, Kuster N. IT’IS database for thermal and electromagnetic parameters of biological tissues, Version 2.6, www.itis.ethz.ch/database.

9. Kraff, O., Fischer, A., Nagel, A.M., Mönninghoff, C. and Ladd, M.E. (2015), MRI at 7 tesla and above: Demonstrated and potential capabilities. J. Magn. Reson. Imaging, 41: 13-33. https://doi.org/10.1002/jmri.24573

10. Christopher J. Lote, (2012), Principles of Renal Physiology, Springer, New York, https://doi.org/10.1007/978-1-4614-3785-7

11. Haneder, Stefan and Konstandin, Simon and Morelli, John N. and Nagel, Armin M. and Zoellner, Frank G. and Schad, Lothar R. and Schoenberg, Stefan O. and Michaely, Henrik J., (2011), Quantitative and Qualitative 23Na MR Imaging of the Human Kidneys at 3 T: Before and after a Water Load, Radiology, 260, 3: 857-865, https://doi.org/10.1148/radiol.11102263

Figures

Fig. 1: Flip angle map for a (a) phantom filled with demineralized water and 35mmol/L ²³Na in all directions and for the whole data set (first column) as well as the inhaled (second column) and exhaled (third column) state and (b) in-vivo measurements. All three respiratory states show a similar field distribution with slight deviations due to breathing as well as other influences as for example SNR. Slices in each direction were chosen to have the kidneys visible (transversal: 38, coronal: 53, saggital: 33, for a 80×80×80 matrix).

Fig. 2: Phantom (a) and in-vivo (b) data in all directions. First column shows the not respiratory sorted data corrected with the unsorted B₁⁺ map, second column shows the inhaled state data corrected with the inhale sorted B₁⁺ map, third and fourth column show the exhaled state data corrected with the exhale sorted B₁⁺ map and the unsorted B₁⁺ map respectively. Slices in each direction were chosen to have the kidneys visible (transversal: 38, coronal: 53, saggital: 33, for a 80×80×80 matrix).

Fig. 3: Quantitative ²³Na concentration in the human kidney at 7T in all directions for not respiratory sorted data, inhaled and exhaled state. Slices in each direction were chosen to have the kidneys visible (transversal: 38, coronal: 53, saggital: 33, for a 80×80×80 matrix).

Tab. 1: Determined concentrations in the whole phantom as well as for the left and right kidney rois from the in-vivo measurement for the different respiratory states (not sorted, inhaled, exhaled) and in comparison for not corrected, B₁⁺ corrected and B₁⁺ as well as B₁^- corrected data.

Tab. 2: Determined concentrations for the left and right kidney for the different respiratory states (not sorted, inhaled, exhaled) and in comparison for no correction, B₁⁺ corrected and B₁⁺ as well as B₁^- corrected data.

Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)

0238