Periodically gapped data spectral velocity estimation in medical ultrasound using spatial and temporal dimensions

Modern medical ultrasound scanners estimate blood velocity distribution by computing the spectrogram of a temporal data sequence, typically using periodogram methods which require long observation windows. Furthermore, an additional B-mode image is often displayed, resulting in gaps in the data at B-mode emissions. We propose a data-adaptive velocity estimator for periodically gapped (PG) data that extends PG-Capon and PG-APES by using two dimensional spatial and temporal data to estimate a one dimensional spectrum. We show through realistic flow simulations that our method improves spectral resolution and reduces leakage in comparison to PG-Capon, PG-APES, and correlogram based gapped data velocity estimators, potentially increasing the maximum detectable velocity and temporal resolution of blood flow using ultrasound.