Spectral velocity estimation in ultrasound using sparse data sets

Abstract

Velocity distributions in blood vessels can be displayed using ultrasound scanners by making a Fourier transform of the received signal and then showing spectra in an $M$ -mode display. It is desired to show a $B$ -mode image for orientation, and data for this have to be acquired interleaved with the flow data. This either halves the effective pulse repetition frequency $f_{\mathrm{prf}}$ or gaps appear in the spectrum from $B$ -mode emissions. This paper presents a technique to maintain the highest possible $f_{\mathrm{prf}}$ and at the same time show a $B$ -mode image. The power spectrum can be calculated from the Fourier transform of the autocorrelation function, and it is shown that the autocorrelation function can be calculated for a sparse set of data where flow and $B$ -mode emissions are interspaced. Both short deterministic sequences of emissions and full random sequences can be used. The dynamic range of the sparse sequence is reduced compared to a full sequence. Typically, a reduction of $20\mathrm{dB}$ is found when using 66% of the data compared to using all data. The theory of the method and examples from simulations of flow in arteries are presented. The audio signal can also be generated from the spectrogram.