Double-Profile Intersection (DoPIo) Ultrasound: Pointwise Shear Elasticity Estimation using Paired Confocal Displacement Profiles

TL;DR

Current ARF-based elastography methods rely on shear-wave propagation, limiting spatial resolution and robustness in complex tissues. DoPIo introduces pointwise elasticity estimation by tracking two confocal displacement profiles with different widths and using their intersection time $t_{\textnormal{int}}$ to infer the shear modulus $G$ via an FEM-derived empirical model. The approach demonstrates accurate modulus estimation in silico and capable discrimination of soft versus stiff regions in vitro and ex vivo, with estimates largely independent of ARF amplitude once a displacement threshold is reached. Limitations include reduced accuracy in stiffer materials and underestimation in viscoelastic tissues like liver, motivating model refinements and data-driven extensions to improve precision and generalizability for quantitative elasticity imaging.

Abstract

Current acoustic radiation force (ARF) based methods for quantifying tissue elasticity primarily rely on shear wave propagation. However, their spatial resolution is limited by the need for spatial averaging, and their accuracy is affected by shear wave guidance, out of plane reflections, and geometric dispersion, which reduce their applicability in mechanically complex tissues. This study introduces a novel technique called Double Profile Intersection (DoPIo) ultrasound, which enables pointwise estimation of shear elastic modulus within the region of ARF excitation by leveraging the scatterer shearing rate. This rate is inferred by tracking ARF induced displacement using two tracking beams with different lateral widths. The wider beam captures scatterers located outside the ARF excitation region that begin to displace as shearing propagates. The time at which the two resulting displacement profiles intersect is mapped to shear elastic modulus using an empirically derived model based on finite element simulations. In silico, DoPIo estimated shear elastic modulus with a median error of -0.02 kPa and a median absolute deviation of 1.98 kPa in elastic materials up to 35 kPa. Experimental validation in vitro and ex vivo demonstrated that DoPIo reliably distinguished softer regions from stiffer ones, and its modulus estimates remained consistent across varying ARF push amplitudes, provided sufficient displacement estimation signal to noise ratio. DoPIo offers a feasible approach for high resolution, on axis shear elasticity estimation and holds promise as a quantitative biomarker that is independent of ARF amplitude.

Figures (8)

• Figure 1: Conceptual illustration of the beam sequence and signal processing employed in DoPIo elastography. From left to right: (1) An acoustic radiation force excitation is applied with a predetermined focal configuration. (2-4) Scatterer displacements are tracked over time using two confocal tracking beams with different lateral widths. (5) The time at which the resulting displacement profiles intersect, denoted as $t_{\textnormal{int}}$, reflects the rate of lateral scatterer shearing and is related to shear elastic modulus through an empirically derived model.
• Figure 2: Flowchart illustrating the DoPIo imaging procedure, including (top row) development of the empirical model and (bottom row) its evaluation using simulated and experimentally acquired displacement datasets. Times-intersect values ($t_{\mathrm{int}}$) are measured at depth $z$ to estimate the shear elastic modulus ($G$).
• Figure 3: Distributions of times-intersect values for simulated materials with varying elasticities, obtained using (a) separate or (b) simultaneous displacement tracking. Box plots show the median, interquartile range, 95% confidence interval, and outliers for displacements measured at the focal depth (25 mm) across 20 independent scatterer realizations. Color and shading denote ARF push and tracking beam combinations as indicated in the legend.
• Figure 4: Statistical comparisons of times-intersect distributions between the simulated materials in Fig. 3, with $p$-values (Kruskal-Wallis test with Bonferroni post-hoc corrections) on the lower triangle of each grid and Hedge's $g$ effect sizes on the upper triangle. Results are shown for the focal configuration pairings achieving the greatest separation between materials for (a) separate tracking (F/1.5 push and F/1.5 and F/3.0 track beams) and (b) simultaneous tracking (F/3.0 push and F/1.5 and F/5.0 track beams).
• Figure 5: Distributions of DoPIo-derived elastic modulus values for simulated materials with varying elasticities, obtained using (a) separate or (b) simultaneous displacement tracking. Box plots indicate the median, interquartile range, 95% confidence interval, and outliers for displacements measured at the focal depth (25 mm) across 20 independent scatterer realizations. Color and shading represent ARF push and tracking beam combinations, as indicated in the legend.