Abstract
3-D ultrasound localization microscopy (ULM) enables comprehensive mapping of microvascular networks by providing micrometer-scale spatial resolution while avoiding projection errors inherent to 2-D ULM imaging. Current 3-D ULM techniques are based on linear pulse sequences combined with spatiotemporal filtering to distinguish microbubble flow from tissue signals. However, singular-value decomposition (SVD)-based filtering demonstrates poor performance in highly mobile organs, suppressing small vessels with slow blood flow along with tissue signals. While imaging based on nonlinear multipulse sequences can isolate microbubble signals regardless of tissue motion, achieving the high-volume acquisition rates required for 3-D ULM remains technically challenging. Here, we present Fast3D-amplitude modulation (AM) imaging, a 3-D nonlinear imaging sequence that achieves a high-volume acquisition rate (225 Hz) using a single 256-channel ultrasound system with a multiplexed 2-D matrix array. We also introduce a motion rejection algorithm that leverages localized microbubble positions to reject respiratory-induced motion artifacts. Fast3D-AM imaging achieved a superior contrast-to-tissue ratio (CTR) than Fast3D, exhibiting a 6.66-dB improvement in phantom studies. In an in vivo rat study, Fast3D-AM demonstrated higher CTR across all SVD cutoffs compared to Fast3D and preserved both major and microvascular structures in whole-organ kidney imaging.