A computational tool for quantitative assessment of peripheral arteries in ultrasound images

Blood velocity vs. time graphic images, that are based on ultrasound Doppler technique, can be used to reveal relationships between data within them and the presence of cardiovascular diseases, among other applications. In clinical protocols, this kind of study involves, typically, hundreds of patients and a manual and tedious segmentation of the blood velocity curve from the images. In this paper we present a computational tool designed to extract quantitative data from these graphics. The algorithm detects the baseline and the spectrum envelope to calculate peak velocity and velocity-time integral (VTI). A comparative analysis between commercial ultrasound systems and the present methodology included measurements of carotid and brachial arteries and echocardiographic exams. The results showed small bias and high correlation for both: systolic peak velocity (bias = 0.02 m/s; r > 0.998; p < 0.001; n = 102) and VTI (bias = 1.25 cm; r > 0.998; p < 0.001; n = 75).