P01.07: Comparison of the volume perfusion index with current 3D power Doppler indexes, an in-vitro study

Objectives: To compare the accuracy and repeatability of the volume perfusion index (VPI), -a mathematical model for calculation of flow developed by our team-, with the current available indexes for threedimensional power Doppler quantification (VI,FI and VFI) in an experimental in-vitro study. Methods: A phantom of flow was designed by using a mechanical pump generating a pulsatile flow with known values and by using elastic artificial vessels and blood mimicking fluid. A set of three and four-dimensional power Doppler volumes was obtained at different flow settings. Three dimensional volumes were quantified by using the VOCAL software included in the machine’s software. Fourdimensional flow measurement was evaluated by using a previously validated algorithm. Repeatability was evaluated through the intraclass correlation coefficient (ICC)for single unpaired tests and the Bland-Altmann (B-A) plot. Accuracy was tested by comparing the actual values against the obtained indexes with a polynomial regression model. Results: Phantom design was successful in generating a pulsatilelike flow and volumes could be taken in 92% of cases. ICC (95% CI) was 0.34 (0.23–0.51) for VI; 0.41 (0.3–0.51) for FI and 0.56 (0.43–0.69) for VFI. For VPI, it was 0.69 (0.48–0.79). B-A plot showed an inter-observer variability (%) (95% CI) of 48 (23–76) for VI; 62 (33–68) for FI; 29 (21–48) for VFI and 22 (15–29) for VPI. r-value (95% CI) was 0.45 (0.33–0.49) for VI; 0.53 (0.38–0.6) for FI; 0.62 (0.51–0.79) and 0.83 (0.65–0.9) for VPI (P < 0.01). Conclusions: Volume Perfusion Index exhibits a more robust internal validity than the traditional three-dimensional power Doppler indexes and might reflect more accurately actual changes in flow.