Answers are grounded in this paper and cite their sources.
Demonstrates enhanced aberration correction in transcranial imaging, suggesting improved imaging quality.
Aberration correction remains a challenge in ultrasound, particularly in transcranial functional ultrasound imaging, where conventional ray-based techniques have shown limited effectiveness due to strong heterogeneity and high sound speed of bone. Wave-Equation Migration Velocity Analysis (WEMVA) solves a similar problem in geophysics, but its implementation is computationally intractable at ultrasound’s MHz frequencies and sub-millimeter grid. We make it tractable by expressing the entire forward, adjoint, and gradient pipeline in a functional-programming paradigm that enables just-in-time-compiled automatic differentiation on a single GPU. Thus, we introduce WEMVA-US, the first full-wave, closed-loop aberration-correction framework translated to medical ultrasound. WEMVA-US propagates the acoustic wave equation in the frequency domain via a Fourier split-step operator and iteratively refines a sound-speed map by minimizing an image-domain loss using an ADAM optimizer. The method does not require external priors and is compatible with any transmit sequence and can correct for aberrations in both soft tissue and transcranial cases in silico, in vitro- and preliminarily in vivo. We demonstrate up to 3× narrowing of the lateral PSF in silico and ∼58% lateral PSF improvement in ex vivo transcranial imaging (macaque skull). WEMVA-US extends to transcranial flow imaging, enhancing SNR and PSF by up to ∼30% in phantoms and ∼10% in vivo (NHPs), laying the groundwork for aberration-corrected real-time transcranial ultrasound imaging.
No takes yet. Share an insight, caveat, or question.
Munot et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: