A Preliminary Study Of Blood-oxygen-level-dependent MRI For The Assessment Of Chronic Kidney Disease

Objective:To explore the clinical application of blood-oxygen-level-dependent MRI for theassessment of chronic kidney disease.Methods:Twelve healthy volunteers were recruited as the control group in this study. Thirty-onepatients with chronic kidney disease, who were enrolled from February2013to September2013, were separated into three group (stage1-2CKD as mild group, stage3-4CKD asmoderate group, stage5CKD as severe group). Regular T1WI and T2WI sequence andmGRE BOLD sequence were performed in Siemens Verio3.0T MR. R2*values ofbilateral cortex and medulla were measured and calculated by two radiologists usingSiemens Syngo software. The differences of cortical and medullary R2*between controlgroup and CKD group were compared with independent-samples t test, while thedifferences between cortical and medullary R2*in each group were analyzed with paired ttest. One-way ANOVA was adopted to compare the cortical and medullary R2*among thecontrol group and three CKD groups. To explore further differences, LSD was performedto compare four groups one another. ROC curve was used to analyze the diagnosticperformance of differences between groups identified by medulla R2*, and the optimumthreshold was determined. P<0.05was considered to indicate a significant difference in alltests.Results:1. On the R2*images, renal cortex could be clearly differentiated from medulla withthe color change form blue to green, yellow till red.2. The medullary R2*was higher than that of renal cortex in both control group and CKD group. There was a significant difference between the cortical and medullary R2*ineach group (control group: t=-20.654, P<0.05; CKD group: t=-38.231, P<0.05). Besides,remarkable differences of cortical and medullary R2*were observed between two groups(cortical R2*: t=-6.649, P<0.05; medullary R2*: t=-5.198, P<0.05).3. There were significant statistical differences of cortical and medullary R2*amongcontrol group, mild CKD group, moderate and severe CKD group (cortical R2*: F=25.849,P<0.05; medullary R2*: F=39.798, P<0.05). Moreover, the R2*of both cortex and medullaascended with the severity of CKD groups. Between the control group and moderate,severe CKD group and among three CKD groups, the differences of cortical R2*weresignificant (P<0.05). While, there was no significant difference of cortical R2*betweencontrol group and mild CKD group (P>0.05). It was also observed that there wereprominent differences of medulla R2*among control group and three CKD groups oneanother.4. Medulla R2*could identify the difference between control group and CKD group,and there was a high diagnostic value with the threshold of33.19s-1. Its sensitivity andspecificity were80.6%and100.0%respectively. Medulla R2*also could identify thedifferences between control group and mild CKD group, mild and moderate CKD groups,moderate and severe CKD groups, in which the diagnostic efficiency of mild and moderateCKD groups was the highest. The threshold was33.81s-1and sensitivity and specificitywere92.9%and80.0%respectively.Conclusions:BOLD-MRI can clearly differentiate normal renal cortex from medulla and it candetect changes of CKD renal oxygenation content, suggesting that it can be used as anoninvasive modality in the assessment of renal function.