Motion during high-framerate ultrasound often blurs microvascular details and leaves in residual tissue error, limiting diagnostic insight. This study introduces a complex-valued gradient optimization that iteratively aligns ultrasound frames, achieving sub-micrometer motion estimation accuracy in simulations, reducing residual motion error 5-fold vs. phase-based and conventional B-mode optimization methods. Applied in vivo to rat kidney ultrasound, it cut residual tissue signal by ~11dB, significantly enhancing microvascular image clarity and promising a more reliable assessment of slow blood flow.
https://www.spiedigitallibrary.org/conference-proceedings-of-spie/13412/134120M/Submicron-tissue-motion-estimation-for-high-framerate-ultrasound-using-complex/10.1117/12.3047260.full
