3922

In vivo measurement of rat brain water content at 9.4T using super-resolution reconstruction

Dennis C. Thomas^1,2, Ana-Maria Oros-Peusquens*¹, Michael Schöneck¹, Antje Willuweit¹, Zaheer Abbas¹, Markus Zimmermann¹, Jörg Felder¹, Avdo Celik¹, and N. Jon Shah^1,3,4,5
¹Institute of Neuroscience and Medicine-4, Forschungszentrum, Jülich, Germany, ²Institute of Neuroradiology, University Hospital Frankfurt, Frankfurt, Germany, ³Institute of Neuroscience and Medicine 11, INM-11, JARA, Forschungszentrum Jülich,, Jülich, Germany, ⁴JARA - BRAIN - Translational Medicine, Aachen, Germany, Aachen, Germany, ⁵RWTH Aachen University, Aachen, Germany

Synopsis

Keywords: Quantitative Imaging, Neurofluids, Super-resolution reconstruction, High-field MRI, ex vivo

Animal models have an indisputable role in the investigation of brain pathology. Investigating water content in vivo could lead to a better understanding of pathogenesis and hence, a robust technique to measure water content using MRI would be very beneficial. Here, we adapt a super-resolution-based technique, previously developed for humans, to the rat brain and report in vivo high-resolution (200μm isotropic) water content maps, obtained using a 9.4T MRI scanner. High resolution, isotropic water content maps of the rat brain are demonstrated. The water content values obtained using the proposed MR technique are compared with ex vivo wet/dry methods.

Introduction

Water content in the mammalian brain is known to be very well regulated¹. Local brain pathologies and systemic diseases have been shown to affect the brain water content^2,3. Since animal models have an irreplaceable role in the study of brain diseases, a robust method to measure water content non-invasively in animals using MRI would allow us to study the changes in animal disease models over time and correlate the measured changes with the ex vivo methods including histology, thus enabling a better understanding of brain diseases. To date, very few studies have been performed to validate animal brain water content obtained by MRI against gold standard ex vivo wet/dry techniques^4-7. Recently, Super-resolution reconstruction (SRR) techniques⁸ were combined with a modified version of the single-scan water content mapping technique⁹ to obtain accurate water content maps of the in vivo human brain at high resolution and isotropic resolution^10,11. In the current work, we modify and adapt this SRR-based technique to perform high and isotropic resolution water content mapping in rats using a small animal 9.4 T MRI scanner. The proposed method is used to measure in vivo brain water content in eight healthy rats. We further validate the obtained water content values with the gold standard ex vivo wet/dry experiments¹² of the excised rat brains.

Methods

Experiments were carried out on a total of eight male Wistar rats (Charles River Laboratories, Sulzfeld, Germany), age range: 10-11 weeks, mean weight = 400 g. The rats were scanned with a 9.4 T MRI scanner¹³ and later sacrificed to carry out the ex vivo studies. Similar to the method described in¹⁰, three orthogonally oriented low-resolution mGRE images (12 echoes) with an in-plane resolution of 200 μm×200 μm and a slice thickness of 600 μm were acquired and recombined using SRR to obtain a single high-resolution (HR) mGRE image (200 μm isotropic resolution) as shown in Fig. 1. The MR parameters were: TR = 3330 ms, FA = 15°, TE0 = 1.35 ms, ∆TE = 2.10 ms, nTE = 12, BW = 650 Hz/pixel, FOV = 38.4 mm x 38.4 mm, matrix size = 192x192, partial Fourier = 6/8, no. of slices = 90, 4 averages per orientation. The total measurement time was 2 hours. Receive field inhomogeneities were corrected for using N4ITK¹⁴. Following the MR experiments, the rats were sacrificed. The rat brain was excised and divided into five regions, namely: frontal left (FL), frontal right (FR), midbrain left (ML), midbrain right (MR) and the cerebellum-brainstem (CB-BS) and ex vivo wet/dry measurements of water content were carried out.

Results

Fig. 2 shows one of the HR in vivo water content maps in all the three orthogonal orientations. Qualitatively, it can be seen that the effect of receive field inhomogeneities is very well corrected. Furthermore, good contrast is achieved between the WM, GM and CSF. The mean whole-brain water content value for this particular rat was 77.43 p.u. At the group level, the whole brain water content was found to be 77.49 +/- 0.33 (mean +/- SD). At a group level, the water content values of all the rat brains measured with the wet/dry method appeared to be similar in both hemispheres but different between the broad regions (Fig. 3). Statistical analysis using repeated measures ANOVA revealed significant differences between the brain regions (p < 0.0001). The water content values of the different regions obtained with MR and ex vivo measurements are shown in Table 1. The SD of the SRR-H2O (0.33 p.u) was found to be comparable with that of the ex vivo measurements (0.31 p.u). Fig. 4A and 4B show the excellent correlation between the water content values obtained using MR measurements and wet/dry measurements, regional: r = 0.902 (p < 0.000001), whole-brain: r = 0.768 (p <0.05).

Discussion and Conclusions

The technique presented here is an adaptation of a recently developed technique in humans using SRR at 3 T. The use of SRR allows one to achieve a better trade-off between SNR and scan time as compared to averaging. The high and isotropic resolution achieved using SRR enables a more accurate delineation of the CSF with minimal partial volume effects thus leading to accurate water content values. Using a TR of 3330 ms and FA of 15° ensures a good contrast between the CSF and brain tissue, as was also demonstrated in¹¹. A very good correlation between the whole brain water content values (Fig 4B, r=0.768) shows a high sensitivity of the MR measurements to changes in whole-brain water content (<1 p.u). The differences in water content values between the broad regions (Fig. 3 and Table 1) are not surprising as different brain regions are known to have different water content values^9,12. These broad differences seen in the reference wet/dry measurements are also accurately reflected in the MR-based water content values, thus validating the accurate correction of the bias field. In conclusion, an in vivo technique to achieve high and isotropic resolution water content maps in rats using SRR has been developed and the accuracy of the technique assessed. The MR derived water content values showed an excellent correlation with the gold standard ex vivo techniques.

Acknowledgements

This work was supported by the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 764513. The authors would like to thank Ms. Kornadt-Beck for help with the animal approval.

References

1. J. E. Hall and A. C. Guyton, Guyton and Hall textbook of medical physiology. 2011.

2. N. J. Shah, H. Neeb, G. Kircheis, P. Engels, D. Häussinger, and K. Zilles, “Quantitative cerebral water content mapping in hepatic encephalopathy,” NeuroImage, vol. 41, no. 3, pp. 706–717, Jul. 2008, doi: 10.1016/j.neuroimage.2008.02.057.

3. C. Laule et al., “Water content and myelin water fraction in multiple sclerosis,” Journal of Neurology, vol. 251, no. 3, pp. 284–293, Mar. 2004, doi: 10.1007/s00415-004-0306-6.

4. W. Lin, R. Venkatesan, K. Gurleyik, Y. Y. He, W. J. Powers, and C. Y. Hsu, “An Absolute Measurement of Brain Water Content Using Magnetic Resonance Imaging in Two Focal Cerebral Ischemic Rat Models,” J Cereb Blood Flow Metab, vol. 20, no. 1, pp. 37–44, Jan. 2000, doi: 10.1097/00004647-200001000-00007.

5. R. Venkatesan et al., “Absolute measurements of water content using magnetic resonance imaging: Preliminary findings in an in vivo focal ischemic rat model,” Magnetic Resonance in Medicine, vol. 43, no. 1, pp. 146–150, 2000, doi: 10.1002

6. A. Schwarcz, Z. Berente, E. Ősz, and T. Dóczi, “In vivo water quantification in mouse brain at 9.4 Tesla in a vasogenic edema model,” Magnetic Resonance in Medicine, vol. 46, no. 6, pp. 1246–1249, 2001, doi: 10.1002/mrm.1324.

7. P. P. Fatouros, A. Marmarou, K. A. Kraft, S. Inao, and F. P. Schwarz, “In Vivo Brain Water Determination by T1 Measurements: Effect of Total Water Content, Hydration Fraction, and Field Strength,” Magnetic Resonance in Medicine, vol. 17, no. 2, pp. 402–413, 1991, doi: 10.1002/mrm.1910170212.

8. H. Greenspan, “Super-Resolution in Medical Imaging,” The Computer Journal, vol. 52, no. 1, pp. 43–63, Feb. 2008, doi: 10.1093/comjnl/bxm075.

9. A.-M. Oros-Peusquens, R. Loução, Z. Abbas, V. Gras, M. Zimmermann, and N. J. Shah, “A Single-Scan, Rapid Whole-Brain Protocol for Quantitative Water Content Mapping With Neurobiological Implications,” Front. Neurol., vol. 10, p. 1333, Dec. 2019, doi: 10.3389/fneur.2019.01333.

10. D. Thomas, A.-M. Oros-Peusquens, D. Poot, and N. J. Shah, “Isotropic water content mapping employing super-resolution reconstruction with acquisition in three orthogonal orientations,” in 2021 ISMRM & SMRT Annual Meeting & Exhibition 15-20 May 2021, Online, May 2021.

11. D. C. Thomas, A.-M. Oros-Peusquens, D. Poot, and N. J. Shah, “Whole-Brain Water Content Mapping Using Super-Resolution Reconstruction with MRI Acquisition in 3 Orthogonal Orientations,” Magnetic Resonance in Medicine, vol. 88, no. 5, pp. 2117–2130, 2022, doi: 10.1002/mrm.29377.

12. M. Schwab, R. Bauer, and U. Zwiener, “The distribution of normal brain water content in Wistar rats and its increase due to ischemia,” Brain Research, vol. 749, no. 1, pp. 82–87, Feb. 1997, doi: 10.1016/S0006-8993(96)01165-1.

13. J. Felder, A. A. Celik, C.-H. Choi, S. Schwan, and N. J. Shah, “9.4 T small animal MRI using clinical components for direct translational studies,” J Transl Med, vol. 15, p. 264, Dec. 2017, doi: 10.1186/s12967-017-1373-7.

14. N. J. Tustison et al., “N4ITK: Improved N3 Bias Correction,” IEEE Trans. Med. Imaging, vol. 29, no. 6, pp. 1310–1320, Jun. 2010, doi: 10.1109/TMI.2010.2046908.

Figures

Figure 1: Schematic showing the steps of data acquisition, image reconstruction and the data processing pipeline. The reconstructed 12-echoes HR mGRE image was used to perform water content mapping using CSF as an internal 100 p.u water reference. The brain masks were generated manually using ITK SNAP. The masked images were then corrected for the receive field inhomogeneities using the N4 Bias field correction algorithm followed by the correction for T2* and finally calibration with the CSF.

Figure 2: Whole-brain water content maps in the three orthogonal orientations from a representative animal: sagittal (top row), coronal (middle row) and axial (bottom row) views. The water content maps can be seen to be well corrected for the bias field which is very strong when using a 4-channel phased array receive coil, as in this study. A good contrast is observed between the WM, GM and CSF in the water content maps.

Figure 3: Box plots of the mean regional water content values (8 rats) of the five brain tissue samples. FL: front left, FR: front right, ML: middle left, MR: middle right, CB-BS: cerebellum-brainstem. Orange line represents the median, the green arrow shows the mean value, the box represents the inter-quartile range, and the whiskers represent the minimum and maximum value of the water content values of the 8 rats. With the ex vivo wet/dry method, the water content values of different broad regions were: front (78.98 +/- 0.40 p.u), middle (77.76 +/- 0.28 p.u) and CB_BS (75.40 +/- 0.48 p.u).

Figure 4: A) Correlation between the water content values obtained with MR measurements and those obtained with wet/dry measurements at a regional level (linear fit: y = 0.70x + 23.39). B) shows the correlation at the whole-brain level (linear fit: y = 0.81x + 14.29). A good correlation in the narrow range of whole brain water content values (77.2 - 78.2 p.u) shows the good sensitivity of the MR-based in vivo water content mapping technique.

Table 1. Comparison of the water content values (mean SD) between the ex vivo (wet/dry) and in vivo (MR) SRR-H2O techniques. FL: front left, FR: front right, ML: middle left, MR: middle right, CB-BS: cerebellum-brainstem

Proc. Intl. Soc. Mag. Reson. Med. 31 (2023)

3922

DOI: https://doi.org/10.58530/2023/3922