DTI And IVIM For The Assessment Of Renal Function Of Diabetes Mellitus By 3.0T MR

Part 1 Diffusion Tensor Imaging for the Assessment of Renal Function of Diabetes MellitusObjectiveTo investigate renal function in diabetes mellitus using magnetic resonance (MR) diffusion tensor imaging (DTI).MethodsThis study was approved by the ethics committee of Shandong Medical Research Institute, and written informed consent was obtained from all participants.20 diabetic and 20 healthy control subjects were recruited for this study. The patients got type II diabetes mellitus but didn’t get clinical diabetic nephropathy and other vascular diseases. The urine protein was negative. DTI data were collected using 3.0T MR scanner and all data were imported into the workstation. The software can handle the data and export the FA and ADC maps and diffusion tensor tractography (DTT). Two experienced radiologists evaluated the quality of FA and ADC maps and measured the cortical and medullary FA and ADC values by selecting multiple regions of interest. The normal distribution test and homogeneity test of variance among the data were conducted by SPSS 20. If the data are normal distribution, t-test was used to test the difference of group comparison and intra-group comparison of cortical and medullary FA and ADC values. Otherwise, Wilcoxon test was used. Values were significant different at P<0.05.ResultsFA values of renal medulla were significantly lower in diabetes mellitus group (0.428±0.015) compared with healthy volunteers(0.454±0.027)(t=-3.785, P<0.05). FA values of renal cortex of diabetes mellitus group (0.243±0.011)were lower than healthy volunteers(0.253±0.020), but it did not show significant difference (t=-1.871, P>0.05). FA values of renal medulla were significantly higher than cortex both in diabetes mellitus group (0.428±0.015,0.243±0.011)and healthy volunteers (0.454±0.027,0.253±0.020) (t=-45.426, P<0.05; t=-26.623, P<0.05). There were significant differences in ADC values of renal medulla (2.194±0.177,2.074±0.127) and cortex (2.488±0.117,2.381±0.164) between diabetes mellitus group and healthy volunteers. ADC values of renal cortex were significantly higher than medulla both i diabetes mellitus group(2.488±0.117,2.194±0.177)and healthy volunteers (2.381±0.164,2.074±0.127) (t= 6.201, P<0.05; t=6.625, P<0.05). FA and ADC maps of healthy volunteers were more clearly visible than those of diabetes mellitus group. Tractography images illustrated the preferred diffusion direction along the tubules and the vessels radial to the renal pelvis. Tractography visualized the reducec radial diffusion tracts in diabetes mellitus group.ConclusionDTI can investigate the renal function of diabetes mellitus before diabetic nephropathy quantitatively. It is helpful for early prevention and treatment.Part 2 Introvoxel Incoherent Motion Imaging for the Assessment of Renal Function of Diabetes MellitusObjectiveTo investigate the value of introvoxel incoherent motion imaging (IVIM) in evaluating renal function of diabetes mellitus.MethodsThis study was approved by the ethics committee of Shandong Medical Research Institute, and written informed consent was obtained from all participants. The patients got type Ⅱ diabetes mellitus but didn’t get clinical diabetic nephropathy and other vascular diseases. The urine protein was negative. IVIM for twenty diabetes mellitus (diabetes mellitus group)and twenty healthy volunteers (control group)of the same age were performed by 3.0T MR scanner. All IVIM data were input into the software of the workstation and processed by bi-exponential model. Two experienced radiologists evaluated the quality of diffusion coefficient (D), perfusion fraction (f), pseudodiffusion coefficient (D*) maps and measured cortical and medullary D, f, D* values by selecting multiple regions of interest by two experienced radiologists. The normal distribution test and homogeneity test of variance were conducted. If the data are normal distribution, t-test was used to test the difference of group comparison and intra-group comparison of cortical and medullary D, f and D* values. Otherwise, Wilcoxon test was used. Values were significant different at P<0.05.ResultsD, f, D* values of renal medulla were significantly different from renal cortex both in diabetes mellitus group and control group(diabetes mellitus group t=9.64, P<0.05; t=12.163,P<0.05; t=15.993, P<0.05; control group t=11.228, P<0.05; t=12.055, P <0.05; t=14.708, P<0.05). D, f values of renal cortex of diabetes mellitus (2.173±0.176,0.293±0.033) were significantly higher than control group (2.069± 0.118,0.269±0.031) (t=2.187, P<0.05; t=2.334, P<0.05). D, f values of renal medulla of diabetes mellitus (1.745±0.091,0.171±0.030) were higher than control group(1.694±0.092,0.166±0.023), but it did not show significant differences(t=1.753, P>0.05; t=0.682, P>0.05). D* values of renal cortex and medulla of diabetes mellitus(127.05± 21.64,29.98±16.38) were higher than healthy volunteers (126.37±26.85,27.14±13.77)without significant differences(t= 0.088, P>0.05; t=0.592, P>0.05). Signal change graph using the bi-exponential model revealed with the increase of b values, the decrease model of signal was non-linear.ConclusionIntrovoxel incoherent motion imaging can investigate the renal function of diabetes mellitus non-invasively and expand its clinical application.