Abstract
Objectives:
To investigate the value of DTI in differentiation of renal allograft rejection from well-functioning stable allograft, using fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values.
Methods:
In this prospective study, 22 transplant recipients with well-functioning stable allograft (group A) and 20 patients with renal allograft rejection (group B + C) were recruited over a period of 19 months from January 2018 to July 2019. DTI-MRI was performed in all the patients, and FA and ADC values were measured in cortical and medullary regions of the transplanted kidney. On biopsy, graft rejection was classified as acute (group B) (n = 7) and chronic graft rejection (group C) (n = 13) based on the BANNF scoring system. Statistical analysis was performed using STATA v.14.0.
Results:
Statistically significant difference between group A and group B + C was noted for cortical (p < 0.001), and medullary (p = 0.003) FA values, and cortical (p = 0.020), and medullary (p = 0.046) ADC values. Cortical(p < 0.001) and Medullary(p = 0.020) FA values showed statistically significant difference between group A and group C, and cortical FA value(p = 0.012) also showed statistically significant difference between group B and group C. AUC (to differentiate between renal allograft rejection and well-functioning stable allograft) for cortical, and medullary FA values and cortical and medullary ADC values were 0.853(p < 0.001), 0.757(p = 0.004), 0.709(p = 0.021) and 0.736(p = 0.009), respectively.
Conclusion and Advances in knowledge:
DTI is a promising functional MRI technique for the non-invasive assessment of renal allograft function. Diffusion parameters, such as FA and ADC values, can be useful in the differentiation of renal allograft rejection from well-functioning stable allograft.
Introduction
Renal transplantation is the treatment of choice for patients with end-stage kidney disease (ESKD). Post-transplant successful functioning of the graft function is the primary concern for both the patients and the treating physician.1 Short-term survival of renal allograft has now increased to more than 90% due to recent advancements in immunosuppressive therapy.2 Renal allograft function is assessed in transplant recipients on a regular basis by urine analysis and serum creatinine levels. Imaging and renal biopsy are performed as and when required.2–4 However, each of them has its limitations. Serum creatinine is an insensitive parameter. It cannot detect early or subclinical graft rejection and slowly developing graft dysfunction. Ultrasonography (USG) is the most commonly used imaging modality. It is useful to evaluate the surgical and vascular complications after renal transplantation, but it is less sensitive to diagnose and monitor graft rejection. Allograft biopsy is considered the gold standard for assessing graft dysfunction; however, it is an invasive procedure with the risk of internal hemorrhage and hematoma formation, and is not a pleasant experience for the patient.2–5 Early diagnosis of renal allograft rejection is still a challenge.6
Magnetic resonance (MR) diffusion tensor imaging (DTI) is a new and advanced non-invasive technique that is studied for the early detection of allograft dysfunction and graft rejection.6,7 Diffusion-weighted imaging (DWI) measures diffusion in three orthogonal planes, and apparent diffusion coefficient (ADC) calculated from diffusion-weighted images, provides mean diffusivity, without any directional information. However, in a healthy kidney, there is a radial orientation of vessels, and renal tubules, which facilitates the diffusion in one direction and restricts in other directions, thus, resulting in anisotropic diffusion. DTI measures diffusion in at least six directions. In addition to ADC, fractional anisotropy (FA) values can also be calculated from DTI data, which provides information on both magnitude of directed diffusion, and main diffusion direction. Microstructural changes in the renal vessels and tubulointerstitial abnormalities seen in graft rejection manifests as altered diffusion parameters, which can be assessed through DTI imaging.8–10
This study aims to investigate the value of DTI in the differentiation of acute and chronic renal allograft rejection from well-functioning stable allograft, using FA and ADC values.
Methods and materials
Study population
This prospective study was approved by the institutional ethics committee and written informed consent was obtained from each renal transplant recipient for inclusion with in the study. The study was conducted collaboratively among departments of radiology, nephrology, and pathology at our institute. Serum creatinine level was measured in all renal allograft recipients. eGFR was calculated using the MDRD equation. Twenty-two consecutive patients with well-functioning stable allograft were included in the study as the control group. Thirty-five consecutive patients with renal allograft dysfunction (eGFR<60 ml/min/1.73 m2) who did not have hypovolemic acute tubular necrosis on clinical grounds were initially recruited in the study. Patients with graft dysfunction were evaluated as per standard protocol with clinical and laboratory evaluation, and ultrasonographic examination. 11 patients with visible structural causes of graft dysfunction on USG, such as hydronephrosis (n = 6), perinephric fluid collection (n = 4), and vascular occlusion (n = 1), were excluded from the study. DTI-MRI was performed in all the patients. Graft biopsy was performed in patients with renal allograft dysfunction, within 6 weeks following MRI. Patients with aetiology other than graft rejection on histological, clinical, and laboratory evaluation were excluded from the analysis, that included three patients with calcineurin inhibitor toxicity, and one patient with BK virus infection. Based on histopathology, patients with allograft rejection were divided into two groups, namely, acute allograft rejection (group B) (n = 7) and chronic allograft rejection (group C) (n = 13). Patients with well-functioning stable allograft were taken as the control group (group A) (n = 22) for comparison (Figure 1).
Figure 1.
Flowchart. (eGFR – estimated glomerular filtration rate; FA – fractional anisotropy; ADC – apparent diffusion coefficient)
DTI-MRI protocol
DTI-MRI was performed on all the renal transplant recipients included in the study, in a 3 T MRI system (3T Ingenia, Philips, The Netherlands). MRI was performed before the biopsy to avoid hindrance in MR image analysis from parenchymal hemorrhage. Noncontrast MRI was performed in the supine position in a 3.0T MR system (Ingenia, Philips, USA). A transverse and coronal T2-weighted half-Fourier rapid acquisition with relaxation enhancement sequence was performed in all patients to exclude visible renal abnormality and to plan the DTI sequence. Coronal fat-suppressed echo-planar imaging sequence was acquired for DTI, with following parameters: number of diffusion direction, 15; b values, 0 and 800 s/mm2 ;TR/TE, 1388 ms/66 ms; FOV, 380 × 299 × 138 mm3; 42 slices; matrix size, 152 × 117; voxel, 2.50 × 2.55 × 3.00 mm; scanning time, 15–20 min. No respiratory compensation was used.
Image analysis
Images were analyzed by an experienced radiologist (Dr. `XX, 17 years of experience), who was blind to clinical and biochemical findings. The images were evaluated on a dedicated PACS workstation (IntelliSpace Portal version 8.0, Philips, The Netherlands). The kidney was divided into three regions, namely, upper pole, middle zone, and lower pole. A Central coronal slice was used to draw the region of interest (ROI). An ROI of >15 mm2 was drawn in both cortex and medulla, in each region of the kidney. Renal sinus and vascular structures were excluded from measurement. A total of six ROIs were drawn per renal allograft (Figure 2). Mean ADC and FA values were noted, and averages were calculated for cortex and medulla.
Figure 2.
Diffusion tensor MRI of pelvic transplant kidney. (a) Coronal T2-weighted image of transplant kidney. (b) ADC map. (c) Fractional anisotropy (FA) map with representative ROI; (d) Color map.
Renal biopsy, and histopathology
In patients with graft dysfunction, graft biopsy was performed percutaneously by an experienced nephrologist, under USG guidance, using an 18G tru-cut biopsy needle, and two cores were taken from the lower pole of the transplanted kidney. Histopathology specimen was evaluated by an experienced renal pathologist, who was blinded to the MR findings. Renal pathologists followed the recent BANFF classification,11 a grading system used for evaluation of renal allograft rejection, and graft dysfunction. Accordingly, renal allograft rejection was classified as acute and chronic allograft rejection (Figure 3).
Figure 3.
Representative histopathological images of acute and chronic graft rejection: (a & b) A representative case of acute cellular and antibody mediated rejection: (a) Note the severe mononuclear inflammation, interstitial oedema and tubulitis. Severe intimal arteritis is also noted in the accurate level artery {black arrow} (Hematoxylin- eosin, 10x magnification) (b) Higher power image to demonstrate tubulitis (Hematoxylin- eosin, 60x magnification). A representative case of chronic antibody mediated rejection: (c) Significant tubulointerstitial chronicity with chronic inflammation, glomerular changes of transplant glomerulopathy with continuing glomerulitis are noted (Masson’s Trichrome, 10x magnification) (d) Note the reduplication of glomerular basement membrane on silver methenamine {black arrowhead} (Jone’s silver methenamine, 20x magnification).
Statistical analysis
All the statistical analysis was performed using STATA software (version 14.0, Texas, USA). Continuous variables including serum creatinine, eGFR, and diffusion parameters (FA and ADC values), were expressed as mean ± standard deviation. Diffusion parameters were compared between well-functioning stable allograft (group A) and renal allograft rejection (group B + C) using unpaired t-test, and between three independent groups, namely well-functioning stable allograft (group A), acute allograft rejection (group B), and chronic allograft rejection (group C) using one-way ANOVA test with Tukey’s post-test. Assumption of homogeneity was found tenable using Levene’s test. Receiver operating characteristic (ROC) analysis was done for FA values using dichotomous group variable, namely, well-functioning stable allograft (group A), and renal allograft rejection (group B + C). Pearson’s correlation coefficients were calculated to analyze the correlation between diffusion parameters, and eGFR. P-value of ≤0.05 was taken as statistically significant.
Results
Demographic and clinical data
A total of 42 renal transplant recipients were included in the study, between January 2018 and July 2019. Age of the study population ranged between 17 years and 53 years, with mean age of 32 years. Out of 42 patients, 37 patients were males and 5 patients were females. Number of weeks after renal transplantation when patients were included in the study ranged between 3 weeks and 834 weeks, with mean of 165.9 weeks. Study population was divided into three groups, namely, well-functioning stable renal allograft (group A) with 22 patients (52.38%), acute graft rejection (group B) with 7 patients (16.67%) and chronic graft rejection (group C) with 13 patients (30.95%). Table 1 shows patient demographics and clinical data.
Table 1.
Demographic and clinical data
| Demographic / clinical data | Well-functioning stable allograft (N = 22) |
Acute graft rejection (N = 7) |
Chronic graft rejection (N = 13) |
|
|---|---|---|---|---|
| Age in years (mean ± SD) | 32.91 ± 9.05 | 32.57 ± 4.42 | 30.15 ± 8.99 | p = 0.641 |
| Sex (N) | M - 19; F - 3 | M - 6; F - 1 | M - 12; F - 1 | - |
| Weeks after renal transplantation (mean ± SD) | 157.32 ± 239.96 | 110.43 ± 99.38 | 210.31 ± 159.27 | p = 0.551 |
| eGFR in ml/min/1.73 m2 (mean ± SD) | 74.73 ± 15.75 | 40.29 ± 22.48 | 33.46 ± 16.51 | p < 0.001 |
F, female; M, male; SD, standard deviation.
Differentiation of renal allograft rejection from the well-functioning stable allograft
Table 2 shows DTI parameters in patients with well-functioning stable allograft(group A) and those with renal allograft rejection (group B + C) Statistically significant differences were found between well-functioning stable allograft and renal allograft rejection, for both cortical and medullary FA and ADC values (p < 0.05). No statistical significance was seen for cortical and medullary ADC values.
Table 2.
Comparing DTI-MRI parameters between well-functioning stable allograft and renal allograft rejection
| DTI parameters (mean ± standard deviation) | Well-function stable allograft (group A) (N = 22) |
Renal allograft rejection (group B + C) (N = 20) |
P-value (A vs (B + C)) |
|---|---|---|---|
| FA_cortex | 0.208 ± 0.032 | 0.164 ± 0.027 | <0.001 |
| FA_medulla | 0.267 ± 0.053 | 0.220 ± 0.041 | 0.003 |
| ADC_cortex | 2.171 ± 0.236 | 1.998 ± 0.226 | 0.020 |
| ADC_medulla | 1.997 ± 0.267 | 1.832 ± 0.249 | 0.046 |
ADC, apparent diffusion coefficient; DTI, diffusion tensor imaging; FA, fractional anisotropy.
We also analysed the diagnostic ability of diffusion parameters to differentiate between renal allograft rejection and well-functioning stable allograft using ROC curves. Area under curve (AUC) for cortical, and medullary FA values, and cortical, and medullary ADC values were 0.853 (p < 0.001), 0.757 (p = 0.004), and 0.709 (p = 0.021), and 0.736 (p = 0.009), respectively, and all were statistically significant.
Figure 4 shows ROC curves of diffusion parameters for differentiation between renal allograft rejection and well-functioning stable allograft.
Figure 4.
Receiver operator (ROC) curves for differentiation between renal allograft rejection and well-functioning stable allograft. (FA – fractional anisotropy; ADC – apparent diffusion coefficient)
Differentiation of acute and chronic allograft rejection from the well-functioning stable allograft
Table 3 shows DTI parameters in patients with well-functioning stable allograft (group A), acute renal allograft rejection (group B), and chronic allograft rejection (group C).
Table 3.
Comparing DTI-MRI parameters between well-functioning stable allograft (group A), acute allograft rejection (group B), and chronic allohraft rejection (group C)
| DTI parameters (mean ± standard deviation) | Group A (N = 22) |
Group B (N = 7) |
Group C (N = 13) |
P-value |
P-value (B vs A) |
P-value (C vs A) |
P-value (B vs C) |
|---|---|---|---|---|---|---|---|
| FA_cortex | 0.208 ± 0.032 | 0.169 ± 0.053 | 0.162 ± 0.029 | <0.001 | 0.012 | <0.001 | 0.897 |
| FA_medulla | 0.267 ± 0.053 | 0.221 ± 0.042 | 0.220 ± 0.042 | 0.012 | 0.084 | 0.020 | 0.998 |
| ADC_cortex | 2.171 ± 0.236 | 2.030 ± 0.221 | 1.981 ± 0.236 | 0.063 | 0.356 | 0.064 | 0.895 |
| ADC_medulla | 1.997 ± 0.267 | 1.884 ± 0.245 | 1.804 ± 0.257 | 0.113 | 0.584 | 0.100 | 0.789 |
ADC, apparent diffusion coefficient; DTI, diffusion tensor imaging; FA, fractional anisotropy.
Cortical and medullary FA values showed a significant difference between the three groups (p ≤ 0.05). Post hoc comparisons to evaluate pairwise differences between groups using the Tukey test were performed, and a significant difference was seen between group C (chronic allograft rejection) and group A (well-functioning stable allograft), for cortical and medullary FA values (p < 0.05). Cortical FA values also showed significant difference between group B (acute allograft rejection) and group A (p = 0.012). No statistically significant difference was seen between group B and group A for ADC and medullary FA values, and between-group C and group A for ADC values (p > 0.05). No statistically significant difference was observed between group B and group C for both ADC and FA values.
Correlation between eGFR and diffusion parameters
Significant Pearson correlation was noted between eGFR, and both FA and ADC value of cortex and medulla. Moderate positive correlation was noted for cortical and medullary FA value (r = 0.68, p < 0.001; and 0.51, p = 0.001, respectively). Weak positive correlation was noted between cortical and medullary ADC value (r = 0.49, p = 0.001; and 0.48, p = 0.001, respectively).12
Discussion
Renal allograft dysfunction is the major concern after renal transplantation, with the risk of reduced graft survival.13,14 The standard procedure for the assessment of renal allograft dysfunction is renal biopsy, which is an invasive procedure with the risk of internal haemorrhage and haematoma formation, and is not a pleasant experience for the patient.10 Nowadays, there is a paradigm shift towards the use of non-invasive functional imaging techniques for the assessment of renal allograft dysfunction, such as blood oxygen level-dependent (BOLD) MRI, diffusion-weighted imaging (DWI), and diffusion tensor imaging (DTI).6 As applied in our study, DTI might be used to evaluate the complex pathophysiology of renal allograft rejection, as it can reflect changes in overall renal diffusivity and renal microstructure.
The diffusion properties of the transplanted kidney, including ADC and FA values, were investigated in various studies. Hueper et al (n = 64) showed a statistically significant relationship between renal allograft function and medullary FA value, while cortical FA value and cortical and medullary ADC value showed no significant correlation.15 A similar study by Palmucci et al (n = 40) showed a significant association between renal transplant function, and medullary FA value, and cortical and medullary ADC value.16 Lanzmann et al (n = 40) showed similar results where statistically significant difference was seen between transplant recipients with eGFR >30 ml/min/1.73m2, and eGFR <30 ml/min/1.73m2, for both cortical and medullary FA and ADC values.7 The above-mentioned studies evaluated diffusion parameters against eGFR or Creatinine clearance as markers of renal function, however, no histopathological correlation of acute or chronic graft rejection was evaluated. Very few studies have evaluated diffusion parameters against histological evidence of graft rejection.
In our study, we evaluated the role of DTI in the differentiation of acute and chronic renal allograft rejection from well-functioning stable allograft, using FA and ADC values. We noticed that cortical and medullary FA values, and ADC values were significantly lower in renal allograft rejection (group B + C) in comparison with well-functioning stable allograft (group A), indicating damage to vascular and tubular microstructure and function in these patients. Also, cortical and medullary FA values were significantly lower in group C (chronic allograft rejection), and cortical FA value was significantly lower in group B (Acute allograft rejection), in comparison with group A. ADC value showed no significant difference between the individual groups. In contrast to our study, Eisenberger et al (n = 15) showed that cortical and medullary ADC values were strongly reduced in patients with acute graft rejection or acute tubular necrosis, in comparison with allografts with stable function.17 However, a smaller sample size of 15 transplant recipients is a major limitation in this study. Consistent with our study, Deger et al (n = 42) showed a statistically significant difference between transplant recipients with stable function, and those with acute graft rejection, for medullary FA value. However, no statistically significant difference was seen for cortical FA value.10 Heterogeneity in the results observed among different studies can be attributed to differences in the MR systems used, study design, acquisition parameters, and workstation used for analysis. We also observed a significant Pearson correlation between eGFR, and both FA and ADC value of cortex and medulla (p < 0.05). Consistent with our study, Palmucci et al, Fan et al and Lanzmann et al, showed a significant correlation between renal function, and medullary FA value, and cortical and medullary ADC value (p < 0.05). However, cortical FA value showed no significant correlation with renal function.7,9,16 The differences in the results are likely due to the smaller sample size of the studies, and differences in the magnetic field strength of the MR scanner (1.5vs 3T), and the study protocol. A study with a larger sample size and standardized protocol may be required to establish the reproducibility of the results.
The primary limitation of our study is the smaller sample size (n = 42). A study with a larger sample size would make it possible to create more homogeneous subgroups of renal transplant patients based on renal allograft age, and renal pathology for more specific results. Histopathological correlation was only available for patients with biochemical renal function impairment, and graft rejection. Renal biopsy correlation was not done for stable allograft recipients as these patients do not require biopsy as per the standard of care. Patients with clinical diagnosis of hypovolemic acute tubular necrosis, BK virus infection induced renal dysfunction and drug induced interstitial nephritis were not included in the study. DTI has a low signal to noise ratio. Therefore, requiring longer acquisition time to compensate for the same, which could be difficult for the patient who need to remain relatively motionless during the scan. This cross-sectional study is a single-center study with DTI-MRI performed in a Philips 3T ingenia MR system and analyzed using a single dedicated PACS workstation. Additional studies are required to test the reproducibility of the results across the different MR systems and workstations. Comparison of DTI with other functional MRI techniques such as BOLD, and arterial spin labeling (ASL), would be helpful to determine whether a single parameter or combination of parameters can best predict renal allograft rejection.
To conclude, DTI is a promising functional MRI technique for the assessment of renal allograft rejection. Diffusion parameters, such as FA, and ADC values, show a significant correlation with graft dysfunction and renal allograft rejection. Being a non-invasive test, DTI-MRI can be used as alternative to renal biopsy for differentiation of graft rejection from well-functioning stable allograft, particularly if repeated tests are required. However, further studies with larger sample size, and standardised protocol are required before clinical application of the same.
Footnotes
Acknowledgment: The authors acknowledge support from the Department of Science and Technology, Science and Engineering Research Board (SB/S3/EECE/206/2016) funding.
Contributor Information
Chandan Jyoti Das, Email: dascj@yahoo.com.
Vijay Kubihal, Email: vijaysk91@gmail.com.
Sambuddha Kumar, Email: 2sambuddha@gmail.com.
Sanjay Kumar Agarwal, Email: skagarwalnephro@gmail.com.
Amit Kumar Dinda, Email: amit_dinda@yahoo.com.
Vishnubhatla Sreenivas, Email: sreenivas_vishnu@yahoo.com.
REFERENCES
- 1. Lodhi SA, Meier-Kriesche H-U. Kidney allograft survival: the long and short of it. Nephrol Dial Transplant 2011; 26: 15–17. doi: 10.1093/ndt/gfq730 [DOI] [PubMed] [Google Scholar]
- 2. Nankivell BJ, Kuypers DRJ. Diagnosis and prevention of chronic kidney allograft loss. Lancet 2011; 378: 1428–37. doi: 10.1016/S0140-6736(11)60699-5 [DOI] [PubMed] [Google Scholar]
- 3. Halawa A. The early diagnosis of acute renal graft dysfunction: a challenge we face. The role of novel biomarkers. Ann Transplant 2011; 16: 90–98. [PubMed] [Google Scholar]
- 4. Hogan JJ, Mocanu M, Berns JS. The native kidney biopsy: update and evidence for best practice. Clin J Am Soc Nephrol 2016; 11: 354–62. doi: 10.2215/CJN.05750515 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Whittier WL, Gashti C, Saltzberg S, Korbet S. Comparison of native and transplant kidney biopsies: diagnostic yield and complications. Clin Kidney J 2018; 11: 616–22. doi: 10.1093/ckj/sfy051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Hollis E, Shehata M, Khalifa F, Abou El-Ghar M, El-Diasty T, El-Baz A. Towards non-invasive diagnostic techniques for early detection of acute renal transplant rejection: a review. The Egyptian Journal of Radiology and Nuclear Medicine 2017; 48: 257–69. doi: 10.1016/j.ejrnm.2016.11.005 [DOI] [Google Scholar]
- 7. Lanzman RS, Ljimani A, Pentang G, Zgoura P, Zenginli H, Kröpil P, et al. Kidney transplant: functional assessment with diffusion-tensor MR imaging at 3T. Radiology 2013; 266: 218–25. doi: 10.1148/radiol.12112522 [DOI] [PubMed] [Google Scholar]
- 8. Morrell GR, Zhang JL, Lee VS. Magnetic resonance imaging of the fibrotic kidney. J Am Soc Nephrol 2017; 28: 2564–70. doi: 10.1681/ASN.2016101089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Fan W, Ren T, Li Q, Zuo P, Long M, Mo C, et al. Assessment of renal allograft function early after transplantation with isotropic resolution diffusion tensor imaging. Eur Radiol 2016; 26: 567–75. doi: 10.1007/s00330-015-3841-x [DOI] [PubMed] [Google Scholar]
- 10. Deger E, Celik A, Dheir H, Turunc V, Yardimci A, Torun M, et al. Rejection evaluation after renal transplantation using Mr diffusion tensor imaging. Acta Radiol 2018; 59: 876–83. doi: 10.1177/0284185117733777 [DOI] [PubMed] [Google Scholar]
- 11. Goldberg RJ, Weng FL, Kandula P. Acute and chronic allograft dysfunction in kidney transplant recipients. Med Clin North Am 2016; 100: 487–503. doi: 10.1016/j.mcna.2016.01.002 [DOI] [PubMed] [Google Scholar]
- 12. Mukaka MM. Statistics corner: A guide to appropriate use of correlation coefficient in medical research. Malawi Med J J Med Assoc Malawi 2012; 24: 69–71. [PMC free article] [PubMed] [Google Scholar]
- 13. Sharif A, Borrows R. Delayed graft function after kidney transplantation: the clinical perspective. Am J Kidney Dis 2013; 62: 150–58. doi: 10.1053/j.ajkd.2012.11.050 [DOI] [PubMed] [Google Scholar]
- 14. Siedlecki A, Irish W, Brennan DC. Delayed graft function in the kidney transplant. Am J Transplant 2011; 11: 2279–96. doi: 10.1111/j.1600-6143.2011.03754.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hueper K, Khalifa AA, Bräsen JH, Vo Chieu VD, Gutberlet M, Wintterle S, et al. Diffusion-Weighted imaging and diffusion tensor imaging detect delayed graft function and correlate with allograft fibrosis in patients early after kidney transplantation. J Magn Reson Imaging 2016; 44: 112–21. doi: 10.1002/jmri.25158 [DOI] [PubMed] [Google Scholar]
- 16. Palmucci S, Cappello G, Attinà G, Foti PV, Siverino ROA, Roccasalva F, et al. Diffusion weighted imaging and diffusion tensor imaging in the evaluation of transplanted kidneys. Eur J Radiol Open 2015; 2: 71–80. doi: 10.1016/j.ejro.2015.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Eisenberger U, Thoeny HC, Binser T, Gugger M, Frey FJ, Boesch C, et al. Evaluation of renal allograft function early after transplantation with diffusion-weighted MR imaging. Eur Radiol 2010; 20: 1374–83. doi: 10.1007/s00330-009-1679-9 [DOI] [PubMed] [Google Scholar]