Human Placental Vasculature Imaging Using Long Ensemble Angular-coherence-based Doppler

Human placental vasculature imaging holds the potential for aiding the scientific investigation on human placental development and enabling early preeclampsia diagnosis. However, the slow flow and small vessels in the placenta require highly sensitive detection techniques. Coherent flow power Doppler (CFPD) shows high sensitivity in hepatic vessel detection. However, for small vessel applications where long ensemble lengths are essential, CFPD is not suitable because its computational cost and data load are prohibitively high. We have recently proven the equivalence of angular coherence of ultrasound to the spatial coherence, and produced B-mode images using a short-lag angular coherence (SLAC) beamformer, which operates on beamsummed data and can be integrated to clinical scanners with far less computational resources. Here we propose and demonstrate the long ensemble angular-coherence-based Doppler(LEAD) imaging technique. Compared to the CFPD technique, LEAD reduces the data load by a factor of 10-15 and the computation intensity by a factor 100-200, and thus enables the detection of small vessels from the angular coherence of the flow signals acquired using long ensemble length Doppler sequences. The LEAD technique is demonstrated in an in vivo human placental vasculature imaging study. Compared with conventional power Doppler (PD) imaging, LEAD improves the signal-to-noise ratio (SNR) by 4-8 dB, which is on par with CFPD. The improvement of SNR translates to reduction of background noise and revelation of the small vessels in human placenta in vivo.