A new technique with high reproducibility to estimate renal oxygenation using BOLD-MRI in chronic kidney disease

作者：Piskunowicz Maciej; Hofmann Lucie; Zuercher Emilie; Bassi Isabelle; Milani Bastien; Stuber Matthias; Narkiewicz Krzysztof; Vogt Bruno; Burnier Michel; Pruijm Menno^*

来源：Magnetic Resonance Imaging, 2015, 33(3): 253-261.

DOI：10.1016/j.mri.2014.12.002

• BOLD-MRI
• Region of interest
• Chronic kidney disease
• Reproducibility
• Imaging analysis
• Furosemide
• Concentric objects
• Onion peel technique

Objectives: To assess inter-observer variability of renal blood oxygenation level-dependent MRI (BOLD-MRI) using a new method of analysis, called the concentric objects (CO) technique, in comparison with the classical ROI (region of interest)- based technique.
Methods: MR imaging (3 T) was performed before and after furosemide in 10 chronic kidney disease (CKD) patients (mean eGFR 43 +/- 24 ml/min/1.73 m(2)) and 10 healthy volunteers (eGFR 101 +/- 28 ml/min1.73 m(2)), and R2* maps were determined on four coronal slices. In the CO-technique, R2* values were based on a semiautomatic procedure that divided each kidney in six equal layers, whereas in the ROI-technique, all circles (ROIs) were placed manually in the cortex and medulla. The mean R2*values as assessed by two independent investigators were compared.
Results: With the CO-technique, inter-observer variability was 0.7%-1.9% across all layers in non-CKD, versus 1.6%-3.8% in CKD. With the ROI-technique, median variability for cortical and medullary 82* values was 3.6 and 6.8% in non-CKD, versus 4.7 and 125% in CKD; similar results were observed after furosemide.
Conclusion: The CO-technique offers a new, investigator-independent, highly reproducible alternative to the ROI-based technique to estimate renal tissue oxygenation in CKD.

• 出版日期2015-4