Standardization of blood flow measurements by automated vascular analysis from power Doppler ultrasound scan

Power Doppler ultrasound imaging provides a non-invasive method to explore tissue vascularity in real-time. It has farreaching clinical utility, for example to assess the perfusion of organs such as the placenta. In this study, a fully automated method was proposed to standardize the power Doppler signals in the placental bed to estimate the perfusion by a technique known as fractional moving blood volume (FMBV). The uterine vasculature was extracted by a region growing algorithm from 3D power Doppler scan and further localized based on the placenta segmentation obtained by a multi-class fully Convolution Neural Network (CNN). The largest vessel close to the volume of interest (VOI) in placenta was identified within which the average power Doppler signal was used as a standardization value to correct the signal attenuation in placenta bed to estimate the FMBV measurement. This is the first successful attempt to automatically identify individual blood vessel segments from a complex uterine vascular plexus. The proposed method was performed in twenty 3D power Doppler scans of first trimester placenta with promising results. The mean ± STD FMBV value was 21.35% ± 9.43%. With further analysis and evaluation in large dataset, the proposed method will serve as an efficient tool for assessing the blood perfusion in placenta bed.