Purpose

Diffusion MRI (DWI) contrast in the kidney involves an interplay of microstructural hindrances and microcirculation (1,2). Two representations are diffusion tensor imaging (DTI)(3,4), which measures directional anisotropy of water motion, and intravoxel incoherent motion (IVIM) (5,6), which separately quantifies microcirculation and microstructural restriction. Some hybrid methods jointly measure flow and structural anisotropy (7-9). Tubular and vascular flow both contribute to apparent perfusion fraction, and some studies have suggested that they can be resolved via rate of pseudodiffusion with multicompartment analysis (10-13). Renal DWI has also been shown to be modulated with cardiac cycle (14-18) and with flow-compensated diffusion gradients (19). These tools provide both means of contrast and opportunities for modeling. We show cardiac gated REnal Flow and Microstructure AnisotroPy (REFMAP) data with and without flow compensation in healthy volunteers to explore its full range of contrast.

Methods

In this HIPAA-compliant and IRB-approved prospective study, 8 volunteers (7 F, 1M ages 22-54) provided written informed consent and had their abdomens imaged in a 3 T MRI system (MAGNETOM Prisma; Siemens Healthcare, Erlangen, Germany) in supine position with posterior spine array and anterior body array RF coils and chest leads for ECG gating. Coronal oblique T2-weighted HASTE images were collected for anatomical reference. Sagittal phase contrast (PC) MRI images through the left renal artery were collected at multiple cardiac phases to estimate pre-systolic, systolic, and diastolic phases for kidney tissue. With a prototype advanced diffusion weighted imaging sequence with dynamic field correction, cardiac triggered oblique coronal DWI (TR/TE 2800/81 ms, matrix 192/192/1, resolution 2.2/2.2/5 mm) were collected at 10 b-values (b=0, 10,30,50,80,120,200,400,600,800 s/mm²) and 12 directions, with both bipolar and flow-compensated diffusion gradients. The same protocols were acquired from a static water phantom. Acquisitions were performed at maximal velocity (systole, 24±6% full cycle), late stable velocity (diastole, 73±5% full cycle), and between trigger point and systole (pre-systole, 9±2% full cycle). Kidneys were individually cropped and mutual information used for motion correction (Firevoxel v356), and further analyzed with custom code (Igor Pro 8.0). REFMAP analysis generated diffusion tensor metrics (λ_i, MD, FA) for the slow tissue diffusivity, scalar perfusion fraction f_p, and mean, axial, and radial pseudodiffusion (D_p) for both cortex and medulla (segmented on b0 and FA maps respectively). Isotropic averaged integrated signal decays vs. b-values were measured in phantom. Velocity waveforms were extracted from the renal artery in PC-MRI data and velocities were extracted from the peak point and at diastole. Independent samples t-tests and one-way ANOVA (SPSS 25.0) evaluated group dependencies on cardiac cycle and flow compensation. Pearson correlation coefficients were measured between left/right averages of diffusion metrics and renal artery velocities, both in absolute terms and as systolic/diastolic ratios.

Results

6 kidneys were omitted from analysis due to poor positioning, throughplane motion, or variable distortion, leaving 10 kidneys in 6 subjects. Figure 1 shows example PC-MRI slice location and velocity waveform in the renal artery and corresponding diffusion metrics and maps in the adjacent left kidney at multiple trigger delays. Figure 2 shows example integrated signal decays for each pulse sequence. Water shows identical monoexponential decay for both sequences, while renal cortex at systole shows higher perfusion fraction for the bipolar case. One-way ANOVA showed all parameters except MD were significantly different between bipolar and flow compensated cases, with the largest effect sizes appearing in FA and f_p (Figure 3). Conversely, only λ₂, MD, and mean/axial/radial D_p showed significant variance with cardiac phase in the global dataset, and with smaller effect sizes than flow compensation. Average D_p in cortex and medulla correlated significantly with arterial velocity. Sys/dias ratios of f_p and FA in cortex and medulla correlated significantly with sys/dias ratios of arterial velocity (Figure 4).

Discussion

The present study indicates that the flow contribution to kidney DWI is affected differently by cardiac cycle and flow compensation. As previously observed, diffusion metrics (particularly those associated with pseudodiffusion) correlate with renal artery blood velocity. The majority contribution to perfusion fraction remains constant with cardiac cycle, with an additional pulsatile component; this may be consistent with tubular and vascular flow respectively. Flow compensation dramatically reduces both compartments but constant and pulsatile fractions remain; this indicates flow compensation filters some flow contributions from both vascular and tubular spaces. The slow compartment shows some cortical pulsatility in MD and significantly reduced medullary FA with flow compensation, suggesting dynamics remains relevant in that compartment. Diffusion time differences between the two waveforms may also play a role.

Conclusion

Cardiac gating and flow compensation induce significant modulation of diffusion and flow anisotropy in the human kidney. Future signal modeling may shed further biophysical light on these contrast variations.

Target audience

Scientists and clinicians interested in diffusion contrast in renal tissue

Acknowledgements

No acknowledgement found.

References

1.Caroli A, Nephrol Dial Transplant 2018; 2. Sigmund EE, Radiology 2012; 3. Ries M JMRI 2001; 4.Basser PJ. NMR biomed 1995; 5.Le Bihan D, Radiology 1986; 6.Muller MF, Eur J Radiol 1998;7.Notohamiprodjo M, MRM 2015;8.Hilbert F, NMR biomed 2016;9.Liu AL JMRI 2018;10. Baalen Sv, JMRI 2017; 11. Periquito JS, QIMS 2021; 12. van der Bel R, Eur J Radiol 2017;13. Stabinska J, MRM 2021;14. Lanzman RS, MRI 2019;15. Wittsack HJ Invest radiol 2012;16. Heusch P, JMRI 2013;17. Ito K, MRI 2018;18. Milani B MRM 2019 19. Wetscherek A,MRM 2015.