Introduction

The imbalance between oxygen demand and oxygen supply is believed to be a cause of several kidney diseases. Blood oxygenation sensitized MRI (T₂^*mapping) can non-invasively provide information about changes in renal oxygenation. Yet, previous experiments combining non-invasive MRI and invasive physiological measurements of the kidney obtained under different (patho)physiological conditions showed that T₂^* does not accurately represent renal tissue oxygenation^1,2. Confounding factors such as tubular volume fraction should be taken in account for the interpretation of renal T₂^* mapping and for a reliable information about renal tissue oxygenation. Diffusion-weighted imaging (DWI) provides a non-invasive method for in-vivo evaluation of tissue water mobility. Water mobility can serve to probe for tissue alterations at a microscopic level and is sensitive to pathophysiological changes in renal tubuli. DWI holds the promise to measure relative changes in tubular volume fraction. Most DWI studies employ Dw-SE EPI or Dw-Spin Echo pulse sequences. Yet, EPI techniques are sensitive to magnetic field inhomogeneities resulting in geometric distortions which are pronounced at ultrahigh magnetic fields (≥7.0 Tesla), while Spin-Echo approaches result in longer acquisition time. To address this issue we propose a diffusion-weighted split-echo Rapid Acquisition Refocusing Enhancement (RARE) variant for DWI of the rat kidney free of geometric distortion. To meet this objective this study compares a diffusion-weighted split-echo RARE pulse sequence in a diffusion phantom, an ex-vivo rat kidney and an in-vivo rat against SE-EPI and traditional Spin-Echo with the goal to assess the feasibility of diffusion-weighted split-echo RARE for renal imaging.

Methods

Stejskal-Tanner preparation was used for diffusion sensitization of RARE³. A pair of diffusion gradients was added before and after the first refocusing pulse. A split-echo acquisition was used to deal with the destructive interferences between even and odd echoes, which is due to unknown phase shifts due to the diffusion sensitization module^4,5. For initial validation a diffusion phantom was built using a 50ml falcon tube filled with a 5% solution of agarose and three substances with known diffusion properties: sunflower oil, de-ionized water and acetone. Phantom experiments were performed to validate b-value and apparent diffusion coefficient (ADC) map calculation by comparing the experimental data with the literature. Ex-vivo experiments using a perfused rat kidney embedded in agarose and in-vivo experiments with an adult female dark Agouti rat with respiration triggering were performed at a 9.4 Tesla small animal scanner (Bruker Biospec, Ettlingen, Germany). Data was reconstructed offline using custom-made MATLAB code. B-values used for ADC validation were: 0, 200, 400, 600 and 800 s/mm². Geometric distortions between Dw-SE EPI and Dw Split-RARE were compared and quantified by a center gravity analysis using Dw-Spin Echo image as a reference⁶. The matrix size and the FOV were adapted to have the same resolution (0.12x0.12x2.0 mm³) in all experiments.

Results

Figure 1.a shows phantom images obtained with diffusion-weighting ranging from b=0 s/mm² to b=800 s/mm² Figure 1.b depicts the corresponding ADC map. ADC values were 0.087, 2.180 and 4.700s/mm² for sunflower oil, water and acetone (Table 1). Figure 2 compares high resolution images using Dw-Spin Echo, Dw-SE EPI and Dw Split-RARE for an axial slices of the diffusion phantom, a coronal slices of an ex-vivo rat kidney and a coronal slice of a rat abdomen in-vivo. Severe geometric distortions are present in the Dw SE-EPI images. Figure 3 compares the geometric distortions between Dw-SE EPI and Dw Split-RARE using Dw-Spin Echo image as a reference. Center of gravity analysis revealed a displacement of the center of gravity in pixels with respect to the Dw-Spin Echo reference of (2.4 ± 0.2) for Dw Split-RARE, due to point spread function broadening, (13.1 ± 6.2) for Dw-SE EPI in the phantom and (2.8 ± 1.0) for Dw Split-RARE and (8.6 ± 3.0) for Dw-SE EPI in ex-vivo.

Discussion and Conclusion

Diffusion-weighting was successfully implemented in Split-echo RARE. ADC measurements in a diffusion phantom were in line with literature values. Unlike Dw-SE EPI, Dw Split-RARE provided high anatomic fidelity. In in-vivo experiments Dw Split-RARE outperformed Dw-SE EPI by showing both kidneys with no geometric distortions and Dw-Spin Echo by being less prone to motion artifacts due to shorter scan time. To conclude, this study demonstrates the feasibility of Dw Split-RARE at ultrahigh fields for renal imaging. Future in-vivo experiments using (ir)reversible test interventions are needed to evaluate the performance of this approach for the detection of changes in the tubular volume fraction.

Acknowledgements

This work was funded by a grant from the German Research Foundation (NI 532/9-1, FOR 1368).