P01.10: Volume perfusion index: mathematical modelling of aortic isthmus flow by four dimensional power Doppler: results from an experimental sheep model

Objectives: Non-invasive and accurate measurement of fetal flow remains an elusive matter. Precision of current tools is very limited and relies mainly in operator skills. The aim of this study was to test a tool for indirect flow measurement, by using power Doppler signal and four-dimensional ultrasound (4DPD) and image data simulation and adjustment tools based on mathematical modelling of particles’ vectorial velocities. Methods: Six near term pregnant sheep were surgically instrumented to access fetal aorta by transventricular catheterization. Actual flow through aortic isthmus was measured directly by a PICCO monitor, as well as by pulsed Doppler. Several sets of 4DPD volumes were taken during the procedure. Measurements were stored in an electronic datasheet. A mathematical algorithm for signal normalization was developed for the project by using Mathematica. A differencial equation matrix was generated and obtained data was evaluated through an error adjustment process. Mean velocity at every time was calculated and compared to actual data. The obtained values were plotted against the calculated values. Repeatability was evaluated by using a Bland Altmann test. Results: A total of 30 sets of measurements were collected. A mean of six measurements by cardiac cycle were digitally obtained from the velocity curve. An n by m matrix of data was designed for every set. A mathematically generated flow curve was plotted with the data built through a multi-layer algorithm. Bland Altmann test showed an inter observer variability (95% CI) of 22.3% (11.5–34.3). The r-value for calculated flow was 0.78 (0.73–0.91) (P < 0.05). Conclusions: 4DPD with mathematical modeling overcomes the limitations of two and three-dimensional moving particles measurement and might be an accurate tool for fetal flow measurement.