Original Article
The role of kidney diffusion tensor magnetic resonance imaging
in children
Mehmet Burak Özkan
a,⇑, Robert Marterer
b, Sebastian Tscheuner
b, Utku Mahir Yildirim
c,
Elif Ozkan
da
Department of Pediatric Radiology, Dr. Sami Ulus Research and Training Hospital, Altindag, Ankara, Turkey
b
Department of Pediatric Radiology, Graz, Austria
cDepartment of Radiology, Medikal Park Hospital, _Izmir, Turkey d
Department of Pediatrics, Sami Ulus Hospital, Ankara, Turkey
a r t i c l e i n f o
Article history: Received 4 April 2016 Accepted 19 August 2016 Available online 28 September 2016 Keywords: Renal DTI Pediatric eGFR Renal insufficiency Radiology
a b s t r a c t
Background: Diffusion tensor imaging (DTI) in magnetic resonance imaging (MRI) provides information about the microstructure of renal tissue and is becoming increasingly useful in the evaluation of relationship between renal structure and function.
Objectives: To investigate, whether DTI allows assessment of renal impairment and pathol-ogy in pediatric patients with decreased renal functions.
Materials and methods: Thirty-two pediatric patients and seventeen healthy children were included in this prospective study. For DTI, a respiratory-triggered coronal echo planar imaging (EPI) sequence was performed. Cortical and medullary mean axial and radial dif-fusivity and fractional anisotropy (FA) were analyzed.
Results: In healthy subjects, the cortical FA values were significantly lower than the medul-lary FA values (p < 0.001). Cortical and medulmedul-lary ADC values showed positive correlations
(r = 0.499, p = 0.041) and a negative correlation with cortical FA values(r = 0.533,
p = 0.028). The eGFR values were negatively correlated with the medullary ADC values
(r = 0.484, p = 0.049) in healthy subjects and positively correlated with thek medullary
values (r = 0.385, p = 0.027). Additionally in the patient group, the age was positively cor-related with FA cortex values (r = 0.411, p = 0.018) and with the medullary ADC values (r = 0.461, p = 0.007). However, the medullary FA values were negatively correlated with
the medullary ADC values (r = 0.363, p = 0.038). Tractography of healthy volunteers
showed a radial arrangement which converged into the pyramids, whereas renal insuffi-ciency patients had irregular arrangement patterns and architectural distortions in the observed areas.
Conclusions: Renal DTI is a promising diagnostic tool in the assessment of microstructural renal changes and correlates with the eGFR. Therefore, it is possible to estimate the arrangement of the tracks which emanate from the renal medulla. Furthermore, diffusion of the water molecules could be carried out. This study demonstrates the usage of DTI in renal pediatric kidneys. Validations in larger cohort groups with histopathological biopsies are needed.
Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by
Elsevier. This is an open access article under the CC BY-NC-ND license (
http://creativecom-mons.org/licenses/by-nc-nd/4.0/).
http://dx.doi.org/10.1016/j.ejrnm.2016.08.015
0378-603X/Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer review under responsibility of The Egyptian Society of Radiology and Nuclear Medicine.
⇑ Corresponding author.
E-mail addresses:burakozkan79@hotmail.com(M.B. Özkan),marterer@gmail.com(R. Marterer),sebastian044@gmail.com(S. Tscheuner),oebler@gmail. com(U.M. Yildirim),eozkan04@gmail.com(E. Ozkan).
Contents lists available atScienceDirect
The Egyptian Journal of Radiology and Nuclear Medicine
1. Introduction
The major function of the kidneys is to balance the water and solute concentration. Kidneys do this physiolog-ically by the help of glomeruli and tubules, which enable filtration and reabsorption of water and solutes [1–4]. Understanding the basic mechanisms of water diffusion in the kidneys could help to improve the understanding of kidney pathologies. Laboratory parameters, such as serum creatinine or estimated glomerular filtration rate (eGFR), are commonly used for the evaluation of renal tubular function[5–7].
Diffusion weighted imaging (DWI) visualizes water motion at the molecular level and provides useful informa-tion on parenchyma microstructure and funcinforma-tion, basically depending on the Brownian motion [5–7]. The apparent diffusion coefficient (ADC) is a quantitative parameter cal-culated from DWI[1]. It was shown that ADC values tend to decrease in chronic kidney diseases, depending upon the degree of renal impairment[5,8].
Kidneys have a well-defined anatomical structure with tubules, collecting ducts and vessels radially oriented toward the pelvis and in which molecules move in a pref-erential direction[9,10]. To investigate molecular diffusion in kidneys, the direction should be evaluated. In healthy kidney vessels, tubules and collecting ducts are involved in water reabsorption and urine formation. These struc-tures have a highly perfectly organized radial arrangement. This kind of arrangement enables motion of water mole-cules to move in different directions, such as right to left or superior to inferior. Therefore, this is the reason for the anisotropic manner of the molecular diffusion in renal parenchyma[9]. This highly perfectly designed structural organization gets damaged in pathological processes. This relationship was demonstrated by previous studies [5,6,9,11].
DTI enables to measure the diffusion of the water mole-cules in at least six vectoral ways. The diffusion anisotropy could be used to understand the microstructure of the renal parenchyma. The fractional anisotropy (FA) describes the degree of anisotropy of a diffusion process whereas ADC describes only the averaged diffusivity in the direc-tional way. Therefore, DTI enables to estimate and to mea-sure the vectoral way of least restricted diffusion[3,12,13]. In the literature, the correlation between age-dependent changes and the ability of DTI to estimate renal tissue injuries due to tumors, pyelonephritis and renal artery stenosis was demonstrated. Additionally, it is known that the renal medulla has higher FA values than the cortex[5,9,14].
Studies in adults have demonstrated the ability to show microstructural changes correlating with functional parameters by utilizing DTI[3,4,9,15–16]. This prospective study’s purpose was to examine the value of renal DTI in children and adolescents.
2. Material and methods
Divided into two cohorts (normal eGFR of P60 ml/ min/1.73 m2, and reduced eGFR between P25 ml/
min/1.73 m2and <60 ml/min/1.73 m2), a total of 50
chil-dren were included in this prospective single center study. A healthy control group consisted of 17 volunteers, the study group of 33 children. The control group consisted of nine males and eight females with a mean age of 8 ± 0.9 years (3.09–14.7 years). The mean age in the study group was 8.5 years ± 0.7 years (3.02–15.3 years), consist-ing of 20 male and 12 female children. One patient had to be excluded from this study group, as corticomedullary differentiation (CMD) had not been reported.
All of them underwent renal DTI, as further described in the renal MRI acquisition section. The duration of the study was one and half year. To be certain about the absence of renal parenchymal disease, volunteer patients additionally underwent a renal ultrasound examination.
Exclusion criteria were MRI contraindications such as ferromagnetic implants or claustrophobia. Because of a possible risk of nephrogenic systemic fibrosis[17], patients with GFRs lower than 25 were not included in the study.
The dimensions and corticomedullar differentiation were controlled before the renal MRI examination. All the values regarding these parameters were within normal values. Regarding the b:0 image values differentiation between the cortex and the medulla could be easily evalu-ated and able to be understood the object regions and edges.
A study group sub-group, representing patients with renal impairment (eGFRP 25 and equal or lower 40), was defined. eGFRs were calculated from serum creatinine measurements for all subjects based on the revised Schwarz formula for children and adolescent [18]. The patients were sedated under the age of eight. They were sedated by local anesthesiological products.
The study was approved by the institutional review board and written consent was obtained from the parents of all children.
3. Renal MRI acquisition
All subjects were scanned in supine position in an Avanto 1.5 T (T) MRI scanner (Siemens Medical Systems, Erlangen, Germany). Spine array (posterior) and body array (anterior) receiver coils were used to maximize image uni-formity. The imaging protocol consisted of a respiratory-gated (axial 2D, 3D) echo-planar fast spin echo sequence (HASTE) and a coronal T2-weighted sequence. After the HASTE sequence, a respiratory-gated, single-shot diffusion tensor imaging – echo planar imaging (DTI-EPI) acquisition was performed on one kidney (b = 0 and 400 s/mm2, 6
directions + null, TR/TE = 2000 ms/75 ms, imaging slice thick-ness = 3 mm, 10 imaging slices/subject). In order to achieve a better resolution and higher SNR values, either the left or right kidney was evaluated. Only single renal evaluation was enough for the determination.
In all children diffusion MR images were acquired prior to the administration of gadolinium. As a routine part of this renal MRI imaging, we had to use the gadolinium administration in order to get the sufficient renal diagnosis except those under eGFR under 25. Although we had to use gadolinium as a part of imaging technique, the renal DTI was evaluated separately in another work station and the
renal MRI determination did not affect the renal DTI results.
4. Imaging post-processing
Image post-processing was performed with Diffusion Toolkit software package (www.trackvis.org, version 0.5).
Tractography was reconstructed using a deterministic streamline algorithm. This algorithm was based on a user-defined region of interest. To be continuous and homogenous within the previous studies, thresholds were set for minimum fractional anisotropy (FA). A region of interest (ROI) analysis of the ADC and FA maps was per-formed to obtain measures of mean cortical and medullary
Fig. 1a. Shows the descriptive characteristics of healthy volunteers.
ADC and FA for each subject. At least three coronal image slices in proximity to the renal hilum were selected for the ROI analysis. Medullary and cortical ROI selections were obtained in the HASTE images with high cor-ticomedullary contrast and high anatomical detail. The ROIs selected in the HASTE images were directly applied to the
cor-responding (and co-registered) ADC and FA maps to obtain cortical and medullary ADC and FA values. Afterward, the mean of all resulting ADC and FA values was calculated. A Dif-fusion Ellipsoid is completely represented by the DifDif-fusion Tensor, D. FA is calculated from the eigenvalues (k k1, k 2, k k3) of the diffusion tensors[1]The eigenvectors (
) give theFig. 2. (a) Shows the correlations of parameters in the study group patients.
directions in which the ellipsoid has major axes, and the cor-responding eigenvalues give the magnitude of the peak in that direction.
5. Statistical analysis
Among both cohorts, mean medullary and cortical dif-fusion parameters (ADC, FA) as well as eGFRs were com-pared using two-tailed Student’s t tests. Mean cortical
and medullary ADC and FA were also plotted as a function of eGFR for all subjects and Pearson’s correlation coeffi-cients (R) were determined. Multivariable regression was used to analyze the dependence of whole-kidney region-of-interest tractography volume on FA, eigenvectors and renal function diagnosis.
As statistical software (SPSS Statistics 21, IBM, USA) was used, P-values lower than 0.05 were assumed to be statis-tically significant.
Fig. 2 (continued)
Table 1a
The DTI results of t-test for the healthy group.
Paired samples test t
Paired differences
Mean Std. deviation Std. error mean 95% confidence interval of the difference
Lower Upper
Pair 1 medullarFA - adcmedulla 1.646212 0.045.584 0.007935 1.662375 1.630049 207.460 Pair 2 facortex - adccortex 1.993091 0.094447 0.016441 2.026580 1.959602 121.226 Pair 3 k1korteks - k1medulla 0.823939 0.137862 0.023999 0.775056 0.872823 34.333 Pair 4 k2korteks - k2medulla 0.197576 0.182962 0.031850 0.132700 0.262451 6.203 Pair 5 k3korteks - k3medulla 0.21393939 0.13665512 0.02378860 0.16548360 0.26239519 8.993 Pair 6 facortex - medullarFA 0.235364 0.023817 0.004146 0.243809 0.226919 56.769
df Sig. (2-tailed) Pair 1 medullarFA - adcmedulla 32 0.000 Pair 2 facortex - adccortex 32 0.000 Pair 3 k1korteks - k1medulla 32 0.000 Pair 4 k2korteks - k2medulla 32 0.000 Pair 5 k3korteks - k3medulla 32 0.000 Pair 6 facortex - medullarFA 32 0.000
6. Results
Comparison between the two groups showed that the FA values in the medulla were significantly lower (p < 0.001) in the study group than in the control group. Additionally, the ADC cortex, ADC medulla, and FA cortex were statistically significant lower values (p = 0.049, 0.001, and 0.001 respectively) in the study group compared to the control group. However thek values did not show statistically significant differences between the cortical values in both groups (p = 0.121). (CompareTable 2a.)
Within the stepwise regression analysis result medullar tract length showed linear regression with ADC of the medulla (p = 0.015) in study group population. Addition-ally medullar FA results showed partial regression with ADC medulla when eGFR was less than 60 in the study group (p = 0.029).
7. Control group
The mean value of the eGFR in the control group was measured to be 84 mL/min (range 72–119 mL/min) (Fig. 1a).
The ADC value was significantly (p < 0.001) higher in the cortex than in the medulla; the FA value was signifi-cantly (p < 0.001) higher in the medulla than in the cortex. Additionally, thek1, 2, 3 medulla values were higher than in the cortex (p < 0.001, 0.0013, and 0.0002 respectively).
In healthy subjects there was negative correlation of eGFR with ADC medulla values (r = 0.484, p = 0.049). For the cortex evaluation, there was negative correlation between ADC cortex values and FA values (r = 0.533, p = 0.028) (Fig. 1b).
8. Study group
In the cortical regions, ADC values were significantly (p < 0.001) higher than in the medulla. In opposite, the FA values were significantly (p < 0.001) higher in the medulla than in the cortex. (Fig. 2a and b) (Tables 1a–1c). Multivariate stepwise regressions in the impaired renal function subgroup (eGFR6 40) showed statistically differ-ing FA values (p = 0.04), when compared to the control group. (The graphics of the detailed analysis are shown inFig. 3.)
Table 1b
The DTI results of t-test for the patient group.
Paired samples test t df Sig. (2-tailed) Paired differences
Mean Std. deviation Std. error mean 95% confidence interval of the difference
Lower Upper
Pair 1 medullarFA - adcmedulla 1.646212 0.045584 0.007935 1.662375 1.630049 207.460 32 0.000 Pair 2 facortex - adccortex 1.993091 0.094447 0.016441 2.026580 1.959602 121.226 32 0.000 Pair 3 k1korteks - k1medulla 0.823939 0.137862 0.023999 0.775056 0.872823 34.333 32 0.000 Pair 4 k2korteks - k2medulla 0.197576 0.182962 0.031850 0.132700 0.262451 6.203 32 0.000 Pair 5 k3korteks - k3medulla 0.21393939 0.13665512 0.02378860 0.16548360 0.26239519 8.993 32 0.000 Pair 6 facortex - medullarFA 0.235364 0.023817 0.004146 0.243809 0.226919 56.769 32 0.000
Table 1c
The DTI correlation values of t-test for the patient group. Correlations
adccortex medullatractlength adcmedulla facortex k1korteks k1medulla k2korteks aage Pearson’s Correlation 0.461
Sig. (2-tailed) 0.007
medullatractlength Pearson’s Correlation 0.527 0.495
Sig. (2-tailed) 0.002 0.003
adcmedulla Pearson’s Correlation 0.527 1 0.345 0.461
Sig. (2-tailed) 0.002 0.049 0.007
k2korteks Pearson’s Correlation 0.461 Sig. (2-tailed) 0.007 k2medulla Pearson’s Correlation 0.392
Sig. (2-tailed) 0.024
k3korteks Pearson’s Correlation 0.476 0.346 0.644
Sig. (2-tailed) 0.005 0.048 0.000
k3medulla Pearson’s Correlation 0.440
Sig. (2-tailed) 0.010
eGFR Pearson’s Correlation 0.385
In contrast to the control group, tractography in the study group did not demonstrate astrict radial arrange-ment of the fibers extending from the proximal portion to the Henle loop. The tracks were loosely arranged along the displaced parenchyma in the study group. This resulted in hollow spaces without tracts in the location of the dilated renal calyces or parts of the renal pelvices, espe-cially in the patients with diagnosed ureteropelvic junction (UPJ) obstruction (Figs. 4a and 4b).
9. Discussion
The main objective of the study was to assess the feasi-bility and potential benefits of diffusion tensor imaging in pediatric kidney diseases. We acquired diffusion MR images of sufficient quality to perform DTI tractography in a cohort of children with normal and abnormal renal functional parameters based on the eGFR values.
The renal parenchyma is divided into cortex and medulla which is totally different in a histological and morphological point of view. Mainly the cortex part is responsible for the ultrafiltration of plasma. This ultrafil-trate moves through the medullary pyramids. Between the medullary pyramids there are also vessels which pass transverse and parallel to the tubules, enabling active secretion and reabsorption of solutes. In the renal pelvic region, the ultrafiltration moves to the renal papillas end-ing into minor and major calyx formend-ing the urine[1,2,9]. The formation of urine and ultrafiltration of the kidney needs a nearly perfectly designed microstructural arrange-ment, in both the cortex and the medulla. In the cortex, there are randomly arranged fibers extending from proxi-mal portions of the glomerule to the distal parts of the con-voluted tubules. The fiber network system in the medulla consisted of a parallel vascular network design. The tubu-lous structures, including the ascending and descending portions of the Henle loop are also found in this parallel designed portion system. In this volumetric design within the comparison of medulla with the cortical region, the fibers found in the medulla were thinner than the cortical tubules[1,9,13,16].
In a previous study with a small population, fibers extending throughout the whole kidney did not show sig-nificant differences between kidneys with normal or abnormal function [9]. Our study’s results support this finding, although whole renal fiber tracts were signifi-cantly higher in the study group than in the control group. This is most likely due to increasing number of fibers with patient age and was not statistically significant in the mul-tivariate regression analysis. Similar findings have been reported in other previous studies[5,6,8,9]. On the other hand, based on the regression analysis, the medullar FA could have a critical role in the determination and estima-tion of the renal parenchymal injury.
In the evaluation of the fractional anisotropy maps, cor-ticomedullary differentiation detected in the study group was fairly poor when compared with the healthy volun-teers. The insufficient resolution of the FA maps between the regions could be as a result of the injury to the loss of fiber tract organization. This could also be due to the
Table 2a The DTI results of t-test between the groups including cortex and medulla. Independent samples test Levene’s test for equality of variances t-test for equality of means F Sig. t d f Sig. (2-taile d) Mean difference Std. err or difference 95% confidence interval of the difference Lower Upper medullarFA Equal variances assumed 2.967 0.0 92 6.927 45 0.000 0.056998 0.008229 0.040424 0.073572 Equal variances not assumed 6.016 20.150 0.000 0.056998 0.009474 0.037244 0.076752 adccortex Equal variances assumed 4.324 0.0 43 2.019 45 0.049 0.050833 0.025178 0.000123 0.101544 Equal variances not assumed 2.532 44.771 0.015 0.050833 0.020076 0.010393 0.091274 adcmedulla Equal variances assumed 0.325 0.571 6.021 45 0.000 0.086000 0.014283 0.057232 0.114768 Equal variances not assumed 5.668 23.811 0.000 0.086000 0.015174 0.054669 0.117331 facortex Equal variances assumed 6.365 0.0 15 7.515 45 0.000 0.022273 0.002964 0.016303 0.028243 Equal variances not assumed 6.363 19.225 0.000 0.022273 0.003500 0.014952 0.029593 k 1korteks Equal variances assumed 2.089 0.155 1.580 45 0.121 0.0 2540 0.01607 0.05777 0.00698 Equal variances not assumed 1.717 34.013 0.095 0.0 2540 0.01479 0.05545 0.00466 k 1medulla Equal variances assumed 0.279 0.600 2.643 45 0.011 0.09578 0.03624 0.02279 0.16878 Equal variances not assumed 2.639 27.378 0.014 0.09578 0.03629 0.02136 0.17020
capacity of detecting DTI in the thinner fibers, which could be related to the Tesla magnitude. On the other hand, Jaimes et al. did a study with a 3T machine reporting sim-ilar results[9].
Regarding the patient group, there is a decrease in diffu-sion anisotropy obtained. This could be caused within the structural parenchymal alterations, such as interstitial fibrosis, tubular atrophy and cellular infiltration, basically involving pathologies the medullar part of the renal sys-tem. Previously particularly including the small group studies presenting the reduction in the medullary frac-tional anisotropy values was reported [4–6,8,9]. In our study group, there were seventeen patients with eGFR less than 40 ml/min/1.73 m2. In this group we identified that
only medullary FA could be a promising diagnostic tool in the assessment of radial fiber arrangement and its corre-lation with eGFR. Based on our study report, we could use the medullary FA as a marker of renal structural integrity alteration, potentially useful index for the diagnosis and outcome of kidney disease.
Hueper et al. [11], hypothesize that the majority of chronic diseases affecting the renal parenchyma probably alter directed diffusion (FA) before free diffusion (ADC). Previous studies mention that ADC and FA are also influ-enced by different several factors as in the renal medulla. In think age of the water molecules and physiology there
could be high alterations during the movement of the transportation [6,9]. ADC is mainly influenced by perfu-sion, whereas the FA is related to the water molecule trans-port in the collecting tubules.
These comments also suggested that DTI is the most appropriate technique for functionally evaluating the kidney as compared to DWI, and the medullary FA repre-sents the most sensitive parameter in detecting renal dam-age in different diseases.
The limitations of the study include a small study pop-ulation of 49 children. There were no patients with an eGFR lower than 25 ml/min/1.73 m2 and only seventeen
with an eGFR between 25 ml/min/1.73 m2 and 40 ml/
min/1.73 m2. We avoided inclusion of patients with a eGFR
lower than 25 ml/min/1.73 m2 because of potentially
harmful MRI contrast agent effects [17]. Due to the restricted study population, age-dependencies in renal dif-fusion parameters were no assessed explicitly. Regarding this, we noticed similar results as in the literature [9]. Additionally, we did not consider or compare different b values in the evaluation of DTI imaging, which possibly could help to optimize acquisition technique. Previous studies demonstrated the effects of different b values [19–21], that influence the FA values regarding the limita-tions for the specific chronic kidney diseases should be evaluated.
Fig. 3. (a–f) Figures are enabled from the regression plot of the test of multiwise step logistic regression test. In this graphics the selected cases represent the study group who is less than or equal to eGFR of 40. The number of selected cases is seventeen kidneys.The unselected cases are the children who have eGFR greater than 40. (a) Shows the relationship of medullar trace length with age. (b) Shows the relationship of ADC medulla with medullar trace length. (c) Shows thepartialregressionplot of medullar FA and age. (d) Shows the scatter plot of medullar FA. (e) Shows the partial regression plot of medullar FA and ADC medulla values. (f) Shows the histogram of selected histogram in the relationship of frequency.
Fig. 3b (continued)
Fig. 3e (continued) Fig. 3d (continued)
10. Conclusions
Our prospective study showed the capability of renal DTI in the usage of alterations induced by various chronic kidney diseases. DTI enables to understand the
renal anisotropy based on physiology without contrast media administration. Medullary FA values seem to be the best promising tool in the association within the complementary of laboratory and morphological parameters.
Fig. 3f (continued)
Fig. 4a. Shows the DTI evaluation of a seven years old boy with eGFR diagnosed with 30 after TrackVis program reformat technique: The long green renal fibers show the extension of from proximal tubules to the Henle Loop. The proximal portion of the fibers getting thinner represents the tubule pathology in medulla region.This image was obtained with a view of 55°.
Conflict of interest
The authors declare that there are no conflict of interests.
References
[1]Zheng Z, Shi H, Zhang J, Zhang Y. Renal water molecular diffusion characteristics in healthy native kidneys: assessment with diffusion tensor MR imaging. PLoS One 2014;9:e113469.
[2]Wang W, Pui MH, Guo Y, et al. MR diffusion tensor imaging of normal kidneys. J Magn Reson Imaging 2014;40:1099–102.
[3]Heusch P, Wittsack HJ, Kröpil P, et al. Impact of blood flow on diffusion coefficients of the human kidney: a time-resolved ECG-triggered diffusion-tensor imaging (DTI) study at 3T. J Magn Reson Imaging 2013;37:233–6.
[4]Lu L, Sedor JR, Gulani V, et al. Use of diffusion tensor MRI to identify early changes in diabetic nephropathy. Am J Nephrol 2011;34:476–82.
[5]Feng Q, Ma Z, Wu J, Fang W. DTI for the assessment of disease stage in patients with glomerulonephritis–correlation with renal histology. Eur Radiol 2015;25:92–8.
[6]Liu Z, Xu Y, Zhang J, et al. Chronic kidney disease: pathological and functional assessment with diffusion tensor imaging at 3 T MR. Eur Radiol 2014;25:652–60.
[7]Gaudiano C, Clementi V, Busato F, et al. Renal diffusion tensor imaging: is it possible to define the tubular pathway? A case report. Magn Reson Imaging 2011;29:1030–3.
[8]Gaudiano C, Clementi V, Busato F, et al. Diffusion tensor imaging and tractography of the kidneys: assessment of chronic parenchymal diseases. Eur Radiol 2013;23:1678–85.
[9]Jaimes C, Darge K, Khrichenko D, et al. Diffusion tensor imaging and tractography of the kidney in children: feasibility and preliminary experience. Pediatr Radiol 2014;44:30–41.
[10] Notohamiprodjo M, Dietrich O, Horger W, et al. Diffusion tensor imaging (DTI) of the kidney at 3 T–feasibility, protocol evaluation and comparison to 1.5 T. Invest Radiol 2010;45:245–54.
[11]Hueper K, Gutberlet M, Rodt T, et al. Diffusion tensor imaging and tractography for assessment of renal allograft dysfunction-initial results. Eur Radiol 2011;21:2427–33.
[12]Fan W, Ren T, Li Q, et al. Assessment of renal allograft function early after transplantation with isotropic resolution diffusion tensor imaging. Eur Radiol 2015;80(3).
[13]Cutajar M, Clayden JD, Clark CA, Gordon I. Test-retest reliability and repeatability of renal diffusion tensor MRI in healthy subjects. Eur J Radiol 2011;80:e263–8.
[14]Seif M, Lu H, Boesch C, et al. Image registration for triggered and non-triggered DTI of the human kidney: reduced variability of diffusion parameter estimation. J Magn Reson Imaging 2014;00:1–8. [15]Feng Q, Ma Z, Wu J, Fang W. DTI for the assessment of disease stage in patients with glomerulonephritis - correlation with renal histology. Eur Radiol 2014;25:92–8.
[16]Notohamiprodjo M, Reiser MF, Sourbron SP. Diffusion and perfusion of the kidney. Eur J Radiol 2010;76:337–47.
[17]Idée JM, Fretellier N, Robic C, Corot C. The role of gadolinium chelates in the mechanism of nephrogenic systemic fibrosis: a critical update. Crit Rev Toxicol 2014;44(10):895–913. Nov. [18]Schwartz GJ, Muñoz A, Schneider MF, Mak RH, Kaskel F, Warady BA,
Furth SL. J Am Soc Nephrol 2009;20(3):629–37. Mar.
Fig. 4b. Shows the anterior-posterior view of the 3-D tractography. The upperportion of the green fibers get thinner and shows loss of connection of network in the radial arrangement at the medullar region. The red fibers shows the crossing of left to right direction. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
[19]Hui ES, Cheung MM, Chan KC, Wu EX. B-value dependence of DTI quantitation and sensitivity in detecting neural tissue changes. Neuroimage 2010;49(3):2366–74.
[20]Chuck NC, Steidle G, Blume I, et al. Diffusion tensor imaging of the kidneys: influence of b-value and number of encoding directions on image quality and diffusion tensor parameters. J Clin Imaging Sci 2013;3:53.
[21] Lee, Vivian S, Manmeen Kaur, Louisa Bokacheva, Qun Chen, Henry Rusinek, Ravi Thakur, Daniel Moses, Carol Nazzaro, Kramer Elissa L. What causes diminished corticomedullary differentiation in renal insufficiency? J Magn Reson Imaging: JMRI 2007;25(4).http://dx. doi.org/10.1002/jmri.2087.
