Multi-Frame Reconstruction via Clutter Filtering for Computational Contrast-Enhanced Ultrasound Microflow Imaging

Existing computational reconstruction methods for plane-wave ultrasound imaging (PWI) often neglect the interference of strong tissue signals with weak flow-related scattering signals in microflow imaging. This study aims to develop a multi-frame computational ultrasound reconstruction technique that integrates multi-frame sparse constraints with RF signal preprocessing strategies to facilitate the suppression of tissue signals and background noise.

We propose multi-frame reconstruction of contrast-enhanced ultrasound microflow imaging. First, to prevent tissue signals from diluting the sparsity of multi-frame data, we apply singular value decomposition directly to the raw radio-frequency data, attenuating tissue-dominated components before beamforming. This yields sparser data and increases the relative contribution of flow-related components in the optimization.

The method was evaluated in contrast-enhanced microflow simulations and an exploratory in vivo experiment involving one rabbit. Under the tested conditions, VrUSi reduced background signals and produced narrower apparent intensity profiles than delay-and-sum and single-frame computational reconstruction. In the single-rabbit experiment, VrUSi yielded higher image-domain BCR values and smaller apparent-profile FWHM values across the analyzed temporal windows. These metrics are interpreted as background suppression and apparent profile compactness under the tested conditions, rather than as evidence of improved physical contrast or spatial resolution. The findings support feasibility only for visualizing flow-related structures in the tested rabbit lower-limb acquisition.