Validation of a Doppler method for peak blood velocity detection in a CFD-simulated carotid model

The peak velocity of the blood flowing in carotid is a parameter of high impact in the diagnosis of cardiovascular diseases. Current measurement methods are based on heuristic thresholds applied to the Doppler spectrum. Unfortunately, they often produce inaccurate assessments due to their sensitivity to noise and to the spectral broadening effect. Recently, a new vector technique has been proposed that solves these shortcomings thanks to a threshold calculated through an accurate mathematical model of the Doppler spectrum. However, the evaluation of its accuracy in real-life conditions is hampered by the difficulty of obtaining a reliable gold standard. Computational fluid dynamics (CFD) simulations, based on real carotid geometries, can help in filling this gap. In this work the proposed peak measurement method is tested on a realistic CFD model of the carotid bifurcation with an eccentric plaque in the internal branch. Common (CCA) and internal (ICA) arteries were separately investigated for a whole cardiac cycle, and the measurements from the method under test were compared to the CFD velocity reference. The obtained errors are +4.3% and +5.2% for the whole heart cycle in CCA and ICA, respectively, and +4.6% and +3.1% for the systolic peak.