Improved ultrasound scatterer localization using Minimum Variance beamforming

Abstract: Super-resolution ultrasound relies on microbubble tracking and determination of their center of mass (COM). Current methods are mainly based on conventionally beamformed images. The Minimum Variance beamformer (MVB) is known to outperform the conventional Delay-and-Sum (DAS) beamformer in terms of lateral resolution. It is therefore hypothesized that the higher resolution yielded by the MVB can lead to more accurate scatterer localization. The MVB calculates a set of data-dependent apodization weights with the objective to preserve the signal from a desired direction while minimizing contributions from other directions. On the other hand, standard beamformers use fixed apodization weights. The MV method was employed to beamform simulated and experimental point scatterer data acquired by a linear array probe. The results show that compared to DAS, the MVB is able to identify a larger number of scatterers in an image and achieves almost 10 times improved lateral Full-Width-Half-Maximum (FWHM), which is the standard resolution metric for individual point scatterers. The latter subsequently improves the COM calculation by a couple of micrometres on average. The combination of the above resolution gains provided by the MVB may be beneficial when reconstructing micro-vessels of less than ten micrometres in diameter.