Renal blood flow quantification in pigs using contrast-enhanced ultrasound: an ex vivo study.

Purpose: The aim of the study was to evaluate a new method for the quantification of renal blood flow using contrast-enhanced ultrasound (CEUS) in an ex vivo pig kidney model. Material and Methods: After approval by the animal ethics committee, 4 pig kidneys were explanted and perfused with Celsior (R) liquid (Imtix Sangstat, Lyon, France) at different flow rates (30, 50, 70 and 90 ml/min) in an ex vivo phantom. A50% diluted solution of SonoVue (R) (Bracco, Milano, Italy) was infused in the artery at 0.5 ml/min. CEUS was performed with an Aplio system (Toshiba, Nasu, Japan) using a broadband linear transducer and pulse subtraction imaging. A total of 152 destruction-reperfusion sequences were acquired and cine loops were digitally stored for further quantification. Three different ROIs were placed upon the anterior, posterior cortex and segmental artery. Signal intensity measurements were performed in linear units and perfusion parameters were automatically extracted using dedicated software. Curve fitting was performed using a monoexponential model in which a time delay parameter was introduced. This fit allowed the assessment of the local blood flow into the region of interest (called "contrast-enhanced blood flow" (CEBF)). The artery mean signal intensity was averaged from the ten frames prior to the destruction phase. The normalized CEBF (nCEBF) was calculated as the ratio between CEBF and the mean arterial signal intensity. The CEBF and nCEBF were compared to the true blood flow indicated by the pump flow rate. Results: The CEBF was correlated to the true blood flow only for the posterior cortical ROI (R-2 = 0.45, p = 0.05). The normalization using arterial signals improved CEBF correlation to true blood flow: nCEBF became correlated to the true blood flow when considering all ROIs (R-2 = 0.94, p < 0.0001) and correlation was improved for both anterior and posterior cortical ROIs (R-2 = 0, 93, p = 0.0004; R-2 = 0, 90, p = 0.0005, respectively). However, a significant kidney-dependent effect was observed for the anterior cortical ROI (p = 0.017) but not for the posterior cortical ROI (p = 0.89). Conclusion: Normalization using arterial signals significantly improved the estimation of blood flow calculated with CEUS.