A new time-domain narrowband velocity estimation technique forDoppler ultrasound flow imaging. I. Theory

摘要

A significant improvement in blood velocity estimation accuracyncan be achieved by simultaneously processing both temporal and spatialninformation obtained from a sample volume. Use of the spatialninformation becomes especially important when the temporal resolution isnlimited. By using a two-dimensional sequence of spatially samplednDoppler signal “snapshots” an improved estimate of thenDoppler correlation matrix can be formed. Processing Doppler data innthis fashion addresses the range-velocity spread nature of thendistributed red blood cell target, leading to a significant reduction innspectral speckle. Principal component spectral analysis of then“snapshot” correlation matrix is shown to lead to a new andnrobust Doppler mode frequency estimator. By processing only the dominantnsubspace of the Doppler correlation matrix, the Cramer-Rao bounds on thenestimation error of target velocity is significantly reduced inncomparison to traditional narrowband blood velocity estimation methodsnand achieves almost the same local accuracy as a wideband estimator. Antime-domain solution is given for the velocity estimate using thenroot-MUSIC algorithm, which makes the new estimator attractive fornreal-time implementation