Comparison of ultrasound shear wave elastography with magnetic resonance elastography and renal microvascular flow in the assessment of chronic renal allograft dysfunction

Abstract

Background

Monitoring of renal allograft function is essential for early identification of dysfunction and improvement of kidney transplant (KTX) outcome.

Purpose

To non-invasively assess renal stiffness in KTX recipients using ultrasound shear wave elastography (USE) in correlation with multifrequency magnetic resonance elastography (MRE), renal allograft function, and renal microvascular flow determined using a novel ultrasound microvascular imaging technique.

Material and Methods

This prospective study investigated 25 KTXs (functional KTX [FCT], n = 14; chronic KTX insufficiency [DYS], n = 11) in 20 KTX recipients (mean age = 43 ± 14 years). USE was performed using a high-frequency broadband linear transducer and compared with MRE. Shear wave velocity (SWV) was correlated with the estimated glomerular filtration rate (eGFR). Qualitative differences in renal microvascular flow were obtained using SMI.

Results

FCT had higher SWV than DYS in both cortex and pyramids (cortex, FCT: 3.75 ± 0.82 m/s vs. DYS: 2.79 ± 0.73 m/s, P = 0.0002; pyramid, FCT: 2.89 ± 0.46 m/s vs. DYS: 2.39 ± 0.34 m/s, P = 0.044). Cutoff values of 3.265 m/s for cortex, 2.535 m/s for pyramids, and 2.985 m/s for combined non-hilar parenchyma provided sensitivities of 72.7%, 77.8%, and 90.9% and specificities of 71.4%, 78.6%, and 85.7% for detecting renal allograft dysfunction with area under the receiver operating characteristic curve (AUC) values of 0.831, 0.841, 0.925 (95% confidence interval [CI] = 0.67–0.99, 0.66–1.02, 0.83–1.03). USE correlated positively with eGFR (r = 0.741, P = 0.0004) and with MRE-derived SWV (r = 0.562, P = 0.004). Renal microvascular flow was decreased in DYS.

Conclusion

USE is sensitive to renal allograft dysfunction, which is characterized by reduced SWV and renal perfusion. USE has higher image resolution than MRE, while MRE has slightly better diagnostic accuracy.

Keywords

Kidney transplantation primary graft dysfunction elasticity imaging techniques ultrasonography magnetic resonance imaging microcirculation

Introduction

Renal allograft failure is characterized histologically by interstitial fibrosis and tubular atrophy (IFTA). IFTA results from different causes including calcineurin inhibitor toxicity and chronic antibody- or T-cell-mediated rejection (1 –3). Monitoring of renal allograft function is essential for early identification of dysfunction and timely initiation of medical treatment and for improved kidney transplant (KTX) outcome. In the clinical setting, renal allograft function is assessed using a multimodal approach including (Doppler) ultrasound examination, protocol biopsy, and analysis of clinical presentation and laboratory results. However, none of these tests alone is specific enough to detect early renal allograft dysfunction. The presence and grade of renal fibrosis can only be determined by renal biopsy, which is controversial due to its invasive nature (4).

Ultrasound shear wave elastography (USE) is a technique that allows non-invasive detection of tissue mechanical properties. Different ultrasound elastography methods are available and most are based either on the analysis of strain in tissue under stress or on the imaging of shear wave propagation in tissue (5). Preliminary studies of KTX were conducted using ultrasound-based elastography methods such as the acoustic radiation force impulse (ARFI) technique (6,7), real-time elastography (8), and transient elastography (9,10). Magnetic resonance elastography (MRE) is another imaging modality capable of detecting changes in the biomechanical properties of the kidney (11). MRE was demonstrated to be sensitive to changes in stiffness associated with chronic dysfunction of KTXs (12,13).

While it is well established that tissue fibrosis alters organ stiffness assessed by various elastography techniques, a number of other influencing factors have emerged more recently (14). Since renal vascularization is high (15), renal blood flow could be an additional contributor to renal stiffness (16).

The first aim of this study was to measure non-invasively renal stiffness in kidney transplant recipients using USE and to correlate USE-derived shear wave velocity (SWV) with renal allograft function and SWV measured by MRE. The second aim was to identify a possible contribution of renal microvascular flow to changes in renal stiffness.

Material and Methods

Participants

Twenty patients (7 women; mean age = 43 ± 14 years; age range = 23–73 years) were prospectively and randomly recruited from KTX recipients referred to our ultrasound center between February and September 2015. The study was approved by the institutional review board and all participants gave written informed consent. Five of the 20 KTX recipients had two renal allografts, maintaining the dysfunctional allograft after a second transplantation, so a total of 25 independent KTXs were investigated.

KTXs were divided into two groups: one with normal renal allograft function (FCT; 14 kidneys); the other with chronic renal allograft insufficiency (DYS; 11 kidneys). Inclusion criteria for DYS were KTX recipients regularly receiving hemodialysis based on clinical confirmation by the department of nephrology of our hospital according to the guidelines (17). Further criteria were an abnormal B-mode appearance such as atrophy and parenchymal thickness <1.5 cm (18) and reduced eGFR of <30 mL/min/1.73m², estimated from blood creatinine levels using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (19) and averaged over the course of four months before the USE study. All the other KTXs were considered as FCT. All renal transplants were located in the iliac fossa and anastomosed to external iliac arteries and veins.

Ultrasound shear wave elastography

USE was performed in the morning by one trained radiologist using a high-end ultrasound device (Aplio500, Toshiba, Otawara, Japan) with a high-frequency broadband linear transducer (5–15 MHz, 10 MHz center frequency). B-mode ultrasound was performed to obtain anatomical information. To minimize the influence of anisotropic effects a standardized protocol was used, which defined the transducer position to be in the middle of the kidney. This protocol allowed us to measure stiffness approximately parallel to Henle loops, collecting tubules, and vascular bundles with shear waves propagating perpendicular to these structures (20). Based on the previous observation that SWV increases with external pressure (21) the transducer was applied using very slight pressure to avoid organ deformation. In each transplant kidney, circular regions of interest (ROIs) (5 mm in diameter) were placed in three different areas (cortex, pyramids, pelvis). Three measurements per region were performed and mean values for each patient were calculated. To improve reliability, ROIs were selected based on shear wave propagation maps reflecting arrival time contour. The real-time display of shear wave propagation helps in identifying the regions in which sufficient dynamic strain for elasticity reconstruction is produced. Good wave quality was defined as parallel and homogeneous wave appearance in the shear wave propagation map. Following “one shot” application, one dataset of one-dimensional SWV was acquired in 5 s. A color-coded elastogram was displayed over the B-mode image (shear wave speed representing the degree of stiffness; red = high stiffness, blue = low stiffness). SWV in units of m/s can be converted to stiffness by |G^*| = ρ₀ c² based on a pure elastic model (where ρ₀ denotes the material density of 1000 kg/m³, an approximation made for biological tissue). For comparison with previously published MRE results (13) (see also the MRE paragraph), mean SWV of pyramidal and cortical regions obtained from USE were calculated and defined as nonhilar renal parenchyma. Due to the superficial location of the KTX and slim physique of our patients (mean BMI = 23.4 ± 2.5 kg/m² in FCT and 22.5 ± 4.5 kg/m² in DYS), we were able to measure the entire KTX within a maximum ROI depth of 4 cm and with patients holding their breath.

Ultrasound microvascular imaging

Ultrasound microvascular imaging (SMI; Superb Microvascular Imaging, Aplio500, Toshiba, Otawara, Japan) is a novel non-invasive processing technique based on Doppler ultrasound that enables visualization of low velocity microvascular flow. The method allows detection of changes in microvascular blood flow, as described for thyroid nodes (22). Similar to conventional Doppler-based examinations, signal quality in SMI is influenced by pulse repetition frequency (PRF), penetration depth, and gain. Therefore, a PRF of 45 pulses/s, a maximum depth of 4.0 cm, and gain adapted to achieve the best signal-to-noise ratio (SNR) were used in all patients. This technique was used for qualitative analysis of randomly selected microvascular flow in the middle third of the renal transplant cortex in all patients.

Magnetic resonance elastography

MRE was performed on a 1.5-T magnetic resonance imaging (MRI) scanner (Magnetom Sonata; Siemens Erlangen, Germany) at 10-Hz increments over a frequency range of 40–70 Hz. For comparison with USE, the magnitude of shear modulus |G^*| derived by MRE was converted to SWV in m/s by c = √(|G^*|/ρ₀) based on a pure elastic model (where ρ₀ denotes the material density of 1000 kg/m³, an approximation made for many biological soft tissues). Unlike USE, MRE, with its low SNR, could not differentiate between pyramidal and cortical regions in dysfunctional KTXs because of advanced parenchymal atrophy (13). We therefore averaged MRE-SWV across these regions, henceforth using the term non-hilar renal parenchyma to denote these regions. Further details of the MRE technique are described in Marticorena Garcia et al. (13). All study patients underwent USE and MRE less than one week apart. Except for patient numbers 8 and 17, all patients investigated in this study were also included in Marticorena Garcia et al. (13).

Statistical analysis

Group mean values were calculated as means with standard errors of the mean. Since all KTX originated from different donors, we treated all KTX, including the ones from the double transplanted patients, as independent organs for statistical analysis. Intra- and intergroup group differences in USE findings were analyzed by Fisher’s least significant difference test. The diagnostic accuracy of USE in identifying patients with dysfunctional KTXs was assessed using area under receiver operating characteristic (ROC) curve (AUC) analysis. The AUC reflects the accuracy of the test in classifying and distinguishing healthy and diseased KTX and tests with AUC > 0.9 and 0.8–0.9 are generally considered excellent and good, respectively. Pearson correlation coefficients were calculated for correlations between USE and renal allograft function as well as between USE and MRE. A P value <0.05 was considered statistically significant. All statistical analysis was performed using GraphPad Prism 6.0 (GraphPad software).

Results

Shear wave elastography was successfully performed in all patients with KTX. Due to advanced renal atrophy, no pyramids were clearly distinguishable in two patients in DYS. Fig. 1 presents examples of ultrasound images of a functional transplant kidney (Fig. 1a) and a dysfunctional transplant kidney (Fig. 1b). In the FCT group, higher SWVs were observed in the cortex compared to pyramids (cortex = 3.75 ± 0.82 m/s vs. pyramid = 2.89 ± 0.46 m/s, P = 0.0004) and the pelvis (cortex = 3.75 ± 0.82 m/s vs. pelvis = 2.46 ± 0.58 m/s, P < 0.0001). In the DYS group, a significant regional difference was only observed between pelvis and cortex (cortex = 2.79 ± 0.73 m/s vs. pelvis = 1.99 ± 0.60 m/s, P = 0.034). Comparison of FCT and DYS groups revealed a significant decrease in SWV in the DYS group in both cortex (FCT = 3.75 ± 0.82 m/s vs. DYS = 2.79 ± 0.73 m/s, P = 0.0002, 26% reduction) and pyramids (FCT = 2.89 ± 0.46 m/s vs. DYS = 2.39 ± 0.34 m/s, P = 0.044, 17% reduction). However, no significant difference was observed in the pelvic region (FCT = 2.46 ± 0.58 m/s vs. DYS = 1.99 ± 0.60 m/s, P = 0.136) (Fig. 2a). The mean relative intra-patient standard deviation was 15% for cortex, 28% for pyramid, and 35% for pelvis.

Fig. 1.

Ultrasound images of (a) a functional transplant kidney and (b) dysfunctional transplant kidney. In both (a) and (b), gray-scale B-mode images of renal cortex, pyramids, and pelvis are shown. Elastograms (red = high stiffness, blue = low stiffness) and shear wave propagation maps are shown in color and displayed in the top and bottom right corners, respectively. Real-time display of shear wave propagation was used to identify regions in which sufficient dynamic strain for elasticity reconstruction was produced. The appearance of homogeneous and parallel wave fronts was interpreted to indicate a successful elastography examination. ROIs of cortex (Co), pyramids (Py), and pelvis (Pe) are encircled by white lines.

Fig. 2.

(a) Bar graphs of shear wave elastography-derived shear wave velocities (SWV) illustrating the 26% lower mean shear wave velocity in cortex and the 17% lower velocity in pyramids in dysfunctional transplant kidneys (cortex, P*** = 0.0002; pyramid, P* = 0.044). Intragroup analysis shows higher SWV in cortex compared to pyramids (P^###= 0.0004) and pelvis (P^####< 0.0001) in functional kidney transplants and cortex vs. pelvis (P^§§= 0.034) in dysfunctional transplants. (b) ROC curve for assessing the utility of renal wave velocity in discriminating functioning and non-functioning transplant kidneys. Co = cortex (red), Py = pyramid (green), Pe = pelvis (black).

To assess the diagnostic performance of USE in detecting KTX dysfunction, cutoffs with the corresponding sensitivity, specificity, and AUC values were determined for the cortex (cutoff = 3.265 m/s; sensitivity =72.7%; specificity = 71.4%; AUC value = 0.831 [95% confidence interval {CI} = 0.67–0.99]; P = 0.005), the pyramids (cutoff = 2.535 m/s; sensitivity = 77.8%; specificity = 78.6%; AUC value = 0.841 [95% CI = 0.66–1.02]; P = 0.007), and the pelvis (cutoff = 2.310 m/s; sensitivity = 72.7%; specificity = 50.0%; AUC value = 0.666 [95% CI = 0.44–0.89]; P = 0.0163). ROC curves for the cortex, pyramids, and pelvis are shown in Fig. 2b. In two patients from the DYS group, differentiation of cortex and pyramids was not possible due to high-grade renal atrophy.

USE-based SWV values in non-hilar parenchymal region were 3.32 ± 0.53 m/s and 2.37 ± 0.72 m/s for FCT and DYS, respectively. The diagnostic accuracy of SWV in the non-hilar parenchyma was higher than achieved by sub-region analysis with 90.9% sensitivity and 85.7% specificity for a cutoff of 2.985 m/s and AUC value of 0.925 (95% CI = 0.83–1.03; P = 0.0003). MRE-SWV was significantly lower (19%) in DYS compared to FCT (FCT = 3.04 ± 2.28 m/s vs. DYS = 2.46 ± 0.25 m/s; P < 0.0001). For MRE, a cutoff of 2.69 m/s resulted in 90.9% sensitivity and 85.7% specificity with an AUC value of 0.945 (95% CI =0.86–1.03; P = 0.0002).

eGFR were 50.3 ± 13.8 mL/min/1.73m² for FCT and 8.5 ± 5.9 mL/min/1.73m² for DYS. In those KTX recipients with two allografts, eGFR was only related to the functional one. USE-derived SWV correlated positively with eGFR (r = 0.741, P = 0.0004). We compared two different elastography modalities and calculated a positive Pearson coefficient (r = 0.562, P = 0.004) for the correlation between USE- and MRE-derived shear wave velocities (Fig. 3).

Fig. 3.

Positive Pearson correlations between USE-derived shear wave velocity and eGFR (a) and MRE-derived SWV (b).

SMI of the cortical microvasculature consistently showed a qualitative decrease in cortical microvascular flow in dysfunctional compared to functional KTXs. Exemplary datasets of DYS and FCT kidneys from two patients are shown in Fig. 4.

Fig. 4.

Ultrasound microvascular imaging (SMI, superb microvascular imaging) demonstrates a decrease in cortical microvascular flow in dysfunctional (b) compared to functional (a) kidney transplants.

The mean interval between kidney transplantation and USE was 65 ± 82 months (FCT = 36 ± 44 months; DYS = 102 ± 106 months). All patients had a normal BMI without significant differences between the two groups of functional and dysfunctional allografts (BMI, FCT = 23.4 ± 2.5 kg/m²; DYS = 22.5 ± 4.5 kg/m², P = 0.597).

Discussion

We examined renal allograft recipients using two elastography imaging modalities, namely USE and MRE. Our findings indicate that USE allows sensitive detection of renal allograft dysfunction. Results of USE and MRE correlate positively with each other. MRE has a slightly better diagnostic accuracy (13) while USE offers higher spatial resolution, allowing examination of renal sub-regions such as cortex, pyramids, and pelvis.

Unlike in liver fibrosis (23,24), USE shows lower renal stiffness in dysfunctional renal transplants with known high-grade renal fibrosis. Since kidneys are highly perfused organs, absorbing 20–25% of the cardiac output (15), we hypothesize that renal stiffness measured by USE includes contributions from fibrosis-related collagen accumulation and perfusion pressure. In patients with transplant dysfunction, loss of interstitial capillaries (2) results in a lower perfusion pressure and hence a decrease in renal stiffness masking the increase in stiffness occurring due to fibrosis. This interpretation is supported by MRI findings using arterial spin labeling (25) and the reduced microvascular signals measured by SMI in DYS compared to FCT (Fig. 4) in our study. In two animal studies using MRE and USE, decreased renal stiffness was found to correlate with diminished renal perfusion (20,26). A recently published study also shows a dependency of renal stiffness due to changes of perfusion pressure in an ex vivo porcine MRE model (27). A decrease in SWV was also observed in native dysfunctional kidneys using ARFI (16,28,29). Similar findings were obtained in renal transplants using real-time elastography (RTE) (8). There are, however, other studies contradicting these findings. In the literature (7, 9, 10, 30), renal stiffness was found to be increased in chronic KTX failure, and no differences in stiffness between functional and dysfunctional renal transplants were observed in Syversveen et al. (6). The discrepancy may be attributable to differences in renal sub-region identification (9,10), transducer force (21), transducer position (31) with respect to anisotropic renal structures (20), and different transducer geometry (convex vs. linear) (31). Furthermore, when comparing elastography results obtained with different technologies, one should account for the specific dynamic range in which the data are accumulated. In general, transient elastography methods test tissue properties in a higher frequency spectrum than time-harmonic elastography using MRE or ultrasound (32). Therefore, our observation of a positive correlation between USE and MRE is encouraging in that it suggests a general agreement of the measured parameter values over a larger dynamic range in renal tissue and that findings are comparable across platforms. In contrast to the cortex and pyramid, no differences between both groups were found in the pelvis. Potentially, due to reduced data quality shown in propagation maps in more heterogeneous, water filled, tubular tissue.

A positive correlation between eGFR and KTX perfusion was shown in a previous study using arterial spin labeling (25). eGFR depends on the amount of fluid filtered through the kidney, and it is moderately correlated with USE (Fig. 3a), underlining the hemodynamic contribution to renal stiffness. This suggests that USE is sensitive to perfusion pressure.

In functional renal transplants, USE-derived SWV differed between cortex and pyramids, which might be due to different microcapillary networks and microcirculation, as cortical and medullary peritubular capillaries are fed by different efferent arterioles (33). Furthermore, it is known that blood flow regulation varies in these two regions (34). Group analysis of functional vs. dysfunctional KTXs shows a larger decrease in cortical stiffness (26%) compared to pyramidal perfusion (17%). This indicates a higher influence of cortical perfusion in dysfunctional grafts, which is in line with the abovementioned differences in microcirculation.

While our results are encouraging, our study has some limitations including the small sample size, lack of blinded analysis and of intra- and inter-observer variability tests, and non-quantitative determination of kidney perfusion. A further technical limitation is the use of circular ROIs which were provided by the scanner software. The complex geometry of cortical tissue would clearly be better captured by adaptive ROIs making the analysis of tissue heterogeneities more robust. Nevertheless, our findings show that USE allows sensitive diagnosis of dysfunctional renal transplants and appears to be a promising tool for non-invasive identification of kidney recipients with allograft dysfunction.

In conclusion, USE has a good diagnostic accuracy for the non-invasive detection of chronic renal transplant dysfunction. USE-derived renal stiffness is higher in functional than in dysfunctional allografts and is associated with a decrease in cortical microvascular blood flow. Renal allograft stiffness correlates positively with glomerular filtration rate. In comparison with MRE, USE has a slightly lower diagnostic accuracy but provides higher spatial resolution.

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Thomas Fischer: Consultant and honoraria (international congress) for Toshiba Medical, Cannon Group. Bernd Hamm: Consultant and honoraria for Toshiba Medical, Cannon Group. The other authors declared no potential conflicts of interest.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Nankivell

Borrows

Fung

et al.

The natural history of chronic allograft nephropathy. N Engl J Med 2003; 349: 2326–2333.

Racusen

Regele

. The pathology of chronic allograft dysfunction. Kidney Int Suppl 2010; (119): S27–32.

Bamoulid

Staeck

Halleck

et al.

Advances in pharmacotherapy to treat kidney transplant rejection. Expert Opin Pharmacother 2015; 16: 1627–1648.

Cecka

. Kidney transplantation in the United States. Clin Transpl 2008, pp. 1–18.

Gennisson

Deffieux

Fink

et al.

Ultrasound elastography: principles and techniques. Diagn Interv Imaging 2013; 94: 487–495.

Syversveen

Brabrand

Midtvedt

et al.

Assessment of renal allograft fibrosis by acoustic radiation force impulse quantification-a pilot study. Transpl Int 2011; 24: 100–105.

Stock

Klein

Cong

et al.

ARFI-based tissue elasticity quantification and kidney graft dysfunction: first clinical experiences. Clin Hemorheol Microcirc 2011; 49: 527–535.

Orlacchio

Chegai

Del Giudice

et al.

Kidney transplant: usefulness of real-time elastography (RTE) in the diagnosis of graft interstitial fibrosis. Ultrasound Med Biol 2014; 40: 2564–2572.

Arndt

Schmidt

Loddenkemper

et al.

Noninvasive evaluation of renal allograft fibrosis by transient elastography-a pilot study. Transpl Int 2010; 23: 871–877.

10.

Nakao

Ushigome

Nakamura

et al.

Evaluation of renal allograft fibrosis by transient elastography (Fibro Scan). Transplant Proc 2015; 47: 640–643.

11.

Muthupillai

Ehman

. Magnetic resonance elastography. Nat Med 1996; 2: 601–603.

12.

Lee

Glockner

Glaser

et al.

MR elastography in renal transplant patients and correlation with renal allograft biopsy: a feasibility study. Acad Radiol 2012; 19: 834–841.

13.

Marticorena Garcia

Fischer

Durr

et al.

Multifrequency magnetic resonance elastography for the assessment of renal allograft function. Invest Radiol 2016; 51: 591–595.

14.

Sigrist

RMS

Liau

Kaffas

et al.

Ultrasound elastography: review of techniques and clinical applications. Theranostics 2017; 7: 1303–1329.

15.

Carlstrom

Wilcox

Arendshorst

. Renal autoregulation in health and disease. Physiol Rev 2015; 95: 405–511.

16.

Asano

Ogata

Tanaka

et al.

Acoustic radiation force impulse elastography of the kidneys: is shear wave velocity affected by tissue fibrosis or renal blood flow?

J Ultrasound Med 2014; 33: 793–801.

17.

Kidney Disease: Improving Global Outcomes Transplant Work G. KDIGO clinical practice guideline for the care of kidney transplant recipients. Am J Transplant 2009; 9(Suppl. 3): S1–155.

18.

Roger

Beale

Cattell

et al.

What is the value of measuring renal parenchymal thickness before renal biopsy?

Clin Radiol 1994; 49: 45–49.

19.

Levey

Stevens

Schmid

et al.

A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150: 604–612.

20.

Gennisson

Grenier

Combe

et al.

Supersonic shear wave elastography of in vivo pig kidney: influence of blood pressure, urinary pressure and tissue anisotropy. Ultrasound Med Biol 2012; 38: 1559–1567.

21.

Syversveen

Midtvedt

Berstad

et al.

Tissue elasticity estimated by acoustic radiation force impulse quantification depends on the applied transducer force: an experimental study in kidney transplant patients. Eur Radiol 2012; 22: 2130–2137.

22.

Machado

Segal

Lyshchik

et al.

A novel microvascular flow technique: initial results in thyroids. Ultrasound Q 2016; 32: 67–74.

23.

Yoon

Lee

Joo

et al.

Hepatic fibrosis: prospective comparison of MR elastography and US shear-wave elastography for evaluation. Radiology 2014; 273: 772–782.

24.

Yin

Glaser

Talwalkar

et al.

Hepatic MR elastography: clinical performance in a series of 1377 consecutive examinations. Radiology 2016; 278: 114–124.

25.

Heusch

Wittsack

Blondin

et al.

Functional evaluation of transplanted kidneys using arterial spin labeling MRI. J Magn Reson Imaging 2014; 40: 84–89.

26.

Warner

Yin

Glaser

et al.

Noninvasive In vivo assessment of renal tissue elasticity during graded renal ischemia using MR elastography. Invest Radiol 2011; 46: 509–514.

27.

Marticorena Garcia SR, Grossmann M, Lang ST, et al. Tomoelastography of the native kidney: Regional variation and physiological effects on in vivo renal stiffness. Magn Reson Med 2017. doi:10.1002/mrm.26892.

28.

Guo

et al.

Acoustic radiation force impulse imaging for noninvasive evaluation of renal parenchyma elasticity: preliminary findings. PLoS One 2013; 8: e68925–e68925.

29.

Bob

Bota

Sporea

et al.

Relationship between the estimated glomerular filtration rate and kidney shear wave speed values assessed by acoustic radiation force impulse elastography: a pilot study. J Ultrasound Med 2015; 34: 649–654.

30.

Grenier

Poulain

Lepreux

et al.

Quantitative elastography of renal transplants using supersonic shear imaging: a pilot study. Eur Radiol 2012; 22: 2138–2146.

31.

Tatar

Teber

Ogur

et al.

Real time sonoelastographic evaluation of renal allografts in correlation with clinical prognostic parameters: comparison of linear and convex transducers according to segmental anatomy. Med Ultrason 2014; 16: 229–235.

32.

Tzschatzsch

Ipek-Ugay

Guo

et al.

In vivo time-harmonic multifrequency elastography of the human liver. Phys Med Biol 2014; 59: 1641–1654.

33.

Mayeux

MacMillan-Crow

. Pharmacological targets in the renal peritubular microenvironment: implications for therapy for sepsis-induced acute kidney injury. Pharmacol Ther 2012; 134: 139–155.

34.

Damkjaer

Vafaee

Moller

et al.

Renal cortical and medullary blood flow responses to altered NO availability in humans. Am J Physiol Regul Integr Comp Physiol 2010; 299: R1449–1455.