Overcoming the Nyquist limit in blood flow velocity estimation

Spectral Doppler ultrasound imaging typically consists of a spectrogram, showing the velocity distribution of the blood, and a brightness (B-) mode image allowing the operator to navigate. It is desirable to have both high spectral and velocity resolution, so that details in the blood flow can be traced, as well as a high B-mode frame rate to allow for tracking of movements and to adjust the position of the transducer. The blood flow signal is often sampled 1) using alternating transmissions for blood flow estimation and for B-mode imaging, or, 2) by acquiring a full Doppler spectrum and then parts of the B-mode image. The former has the disadvantage that it halves the sampling rate, making it likely that aliasing will occur when imaging fast moving blood or deeply positioned vessels; the latter that gaps appears in the spectrogram, and that if the frame rate of the B-mode images is slow, it will be difficult to track movements. Adaptive methods have been implemented to circumvent such problems, but even so, to get an acceptable frame rate of the B-mode images, the number of transmissions for Doppler estimation will be limited, restricting the spectral resolution. Alternatively, one may use an irregularly spaced emission pattern, but existing work on the topic is limited and generally suffers from poor resolution and spurious velocity components resulting from the irregular sampling pattern. In this paper, we examine the BIAA algorithm, showing that this approach allows for an accurate velocity estimate even from irregularly sampled measurements. Using an irregular emission pattern, with half the emissions used to form the B-mode image, the remaining emissions are found to yield accurate velocity estimates without reducing the maximally measurable velocity and without the spurious velocity components. Moreover, we show that the approach will allow for the same maximal velocity without aliasing as if all emissions would have been used for the velocity estimation.