4984

Kidney stone discrimination using ultrashort TE MRI

Johannes Fischer¹, Agazi Tesfai¹, Ute Ludwig¹, Philippe Fabian Müller², Arkadiusz Miernik², and Michael Bock¹

¹Department of Radiology, Medical Physics, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany, ²Department of Urology, University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany

Synopsis

Kidney stone disease (urolithiasis) is not only very painful, but can also pose serious health risks, when the fragmentation of infected kidney stones releases bacteria, that may cause post-operative sepsis. In this work we show the ability of Magnetic Resonance Imaging (MRI) to discriminate between common types of kidney stones using relative signal intensity and T₂* relaxation times.

Introduction

In the US, the number of reported cases of kidney stone disease is increasing, and a prevalence of as high as 8.8% has been reported¹. Stone formation is linked with obesity, low fluid intake and diabetes². The role of bacteria in stone formation is subject of current research³. While small stones can pass the ureter spontaneously, larger stones require fragmentation (lithotripsy) by focused ultrasound or laser light. During fragmentation bacteria trapped in the kidney stones can be released, which might cause severe post-operative sepsis⁴. Prior knowledge of the kidney stone composition may help to identify stones prone to infection, so that the clinical treatment can be adapted. Previous work has shown that different types of kidney stones can be discriminated by Computed Tomography (CT)^5,6, whereas MRI has shown no significant differences between stone types⁷. In this work we use UTE MRI and T₁ and T₂* relaxometry to discriminate between ex-vivo carbonapatite and uric acid kidney stones, the first of which has been shown to be prone to bacterial infection⁸.

Materials and Methods

For relaxometric measurements, both kidney stones (Fig. 1) were placed in saline solution and imaged on a clinical 3T MR system (Prisma Fit, Siemens Healthineers; Erlangen, Germany) using a 3D UTE sequence. To measure T₂*, in total 8 images were acquired at increasing echo times (TE=50,80,150,300,600,1000,2000,4000 $\mu$ s at $\alpha$ =16°, TR=7.6 ms), and for T₁ quantification 8 data sets with different excitation angles ( $\alpha$ =5,10,15,20,25,30,40,50 ° at TE=50 $\mu$ s, TR=2.94 ms) were measured. For each image, 60.000 radial spokes were acquired and reconstructed in a 160x160x160 matrix resulting in a spatial resolution of 0.75 mm, bandwidth was 2000 Hz/pixel for all 16 measurements. For the T₁ measurements 3 averages were acquired to increase SNR. The samples were placed on the axis of a custom build Tx/Rx solenoid coil (Fig. 2). Transmitter adjustments were performed using the saline water in the sample.

A voxel-wise fit of relaxation parameters T₁ and T₂* was performed for all voxel containing kidney stone material. For the analysis of T₂*, a mono-exponential fit was first performed on the image data $S$ . For a coefficient of determination $R^2$ below 0.99, a second, bi-exponential model was used. For T₁ values, the FLASH-equation⁹ was fit to the data:

$S(\mathrm{TE})=a\mathrm e^{-\frac{\mathrm{TE}}{{T^*_2}_a}}$

$\text{or, if }R^2< 0.99\quad S(\mathrm{TE})=a\mathrm e^{-\frac{\mathrm{TE}}{{T^*_2}_a}}+b\mathrm e^{-\frac{\mathrm{TE}}{{T^*_2}_b}}$

$S(\alpha)=M_0\mathrm{sin}\alpha\frac{1-\mathrm e^{-\frac{\mathrm{TR}}{T_1}}}{1-\mathrm{cos}\alpha\,\mathrm e^{-\frac{\mathrm{TR}}{T_1}}}$

Here, $a$ and $b$ describe the relative amplitudes of the short and long transverse relaxation components ${T^*_2}_a$ and ${T^*_2}_b$ . $M_0$ is the fully relaxed longitudinal magnetization and TR the repetition time of the UTE sequence.

Results

The MR images in figure 3 show a hyperintense signal compared to the surrounding saline solution in the carbonapatite sample whereas the uric acid stone appears hypointense compared to the saline solution. Results from the fit of the relaxation parameters to the signal in a single voxel are exemplified in figure 4 and show a strong difference in the short T₂* relaxation time. Further analysis of the short T₂* fit parameter shows a strong difference between both samples, as can be seen from the scatterplot in figure 5. Only voxels with a significant signal relaxation of both relaxation times - where the amplitude

$a$ of the short T₂* component is higher than

$b/5$ - are considered in order to suppress erratic fit results due to noise in images acquired with short TE.

Conclusion

The difference in image intensity alone allows for a separation of carbonapatite and uric acid stone with MR imaging, which encourages further investigation of other kidney stone materials, such as calcium oxalate. Additionally, both samples also differ in their short T₂* component, whereas the long T₂* component in both is largely similar.

Acknowledgements

Grant support from the Deutsche Forschungsgemeinschaft (DFG) under grant numbers BO 3025/8-1 and UL 1187/6-1 is gratefully acknowledged.

References

[1] C. D. Scales, A. C. Smith, J. M. Hanley, und C. S. Saigal, „Prevalence of Kidney Stones in the United States“, Eur. Urol., Bd. 62, Nr. 1, S. 160–165, Juli 2012.

[2] S. R. Khan u. a., „Kidney stones“, Nat. Rev. Dis. Primer, Bd. 2, S. 16008, Feb. 2016.

[3] A. L. Schwaderer und A. J. Wolfe, „The association between bacteria and urinary stones“, Ann. Transl. Med., Bd. 5, Nr. 2, Jan. 2017.

[4] O. Shoshany u. a., „Percutaneous nephrolithotomy for infection stones: what is the risk for postoperative sepsis? A retrospective cohort study“, Urolithiasis, Bd. 43, Nr. 3, S. 237–242, Juni 2015.

[5] D. E. Zilberman, M. N. Ferrandino, G. M. Preminger, E. K. Paulson, M. E. Lipkin, und D. T. Boll, „In Vivo Determination of Urinary Stone Composition Using Dual Energy Computerized Tomography With Advanced Post-Acquisition Processing“, J. Urol., Bd. 184, Nr. 6, S. 2354–2359, Dez. 2010.

[6] K. Wilhelm u. a., „Focused Dual-energy CT Maintains Diagnostic and Compositional Accuracy for Urolithiasis Using Ultralow-dose Noncontrast CT“, Urology, Bd. 86, Nr. 6, S. 1097–1103, Dez. 2015.

[7] E.-S. H. Ibrahim u. a., „Detection of different kidney stone types: an ex vivo comparison of ultrashort echo time MRI to reference standard CT“, Clin. Imaging, Bd. 40, Nr. 1, S. 90–95, Jan. 2016.

[8] H. Ansari, A. A. Sepahi, und M. A. Sepahi, „Different Approaches to Detect “Nanobacteria” in Patients with Kidney Stones: an Infectious Cause or a Subset of Life?“, Urol. J., Bd. 14, Nr. 5, S. 5001–5007, Aug. 2017.

[9] A. Haase, J. Frahm, D. Matthaei, W. Hanicke, und K.-D. Merboldt, „FLASH imaging. Rapid NMR imaging using low flip-angle pulses“, J. Magn. Reson. 1969, Bd. 67, Nr. 2, S. 258–266, Apr. 1986.

Figures

Figure 1: Photograph of the carbonapatite (left) and uric acid (right) samples in saline solution.

Figure 2: Photograph of the experimental setup. The Falcon tubes with the kidney stones in saline solution (top) are placed on the axis of a custom build solenoid Tx/Rx coil (bottom).

Figure 3: MR images of the samples shown in Figure 1, acquired at TE=50

$\mu$ s and a=16° during the same acquisition. The carbonapatite sample (left) appears hyperintense compared to the surrounding saline solution, the uric acid sample (right) appears hypointense.

Figure 4: Evolution of carbonapatite (blue dots, a) and uric acid (red crosses, b) signal over echo time and excitation angle, as well as fits to a bi-exponential decay and the signal equation (green) for an arbitrary voxel in each sample.

Figure 5: Scatterplot showing the distribution of T₂* over T₁ values in each voxel containing either one of the kidney stone samples (carbonapatite in blue, uric acid in red). The graph on the left shows the short T₂* values obtained with the bi-exponential fit, the long T₂* components of the bi-exponential fit are shown on the right. The T₂* values from the mono-exponential fit only assume values above 500

$\mu$ s and therefore appear only in the right axis.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)

4984