Improved Ultrasound Imaging Performance With Complex Cumulant Analysis.

Abstract

Ultrasound localization microscopy (ULM) breaks the acoustic diffraction limit. However, the temporal resolution of ULM is relatively low because of a long data-acquisition time. Inspired by super-resolution optical fluctuation imaging (SOFI), in this paper, we propose a method for ultrasound imaging with improved imaging performance, which is achieved by using cumulant analysis. Specifically, to eliminate the axial oscillations, here, the cumulant analysis framework is extended, which is used to process the complex-valued analytic signals rather than the real-valued signals. The results from the numerical simulations and in vitro physical phantom experiments indicate that by generalizing cumulant analysis to complex-valued signals, a high imaging performance is achieved with an improvement of ∼35%-42% (lateral direction) and ∼41%-42% (axial direction) in the resolution compared with the temporal mean envelope image, in terms of full-width-half-max (FWHM). In particularly, the axial oscillations appearing in the real cumulant images are effectively eliminated by the complex cumulant analysis. Moreover, the proposed method can easily take advantage of SOFI. In the phantom experiment, a short data-acquisition time (∼2 sec) is enough to obtain the improved spatial resolution. The proposed method offers an implementation of US with high spatial resolution, fast data-acquisition speed, and axial oscillations removal characteristics. The method provides the potential in US imaging fast biological processes in vivo.