Robust Tracking of Small Displacements With a Bayesian Estimator

Area of Science:

• Medical imaging
• Biomedical engineering
• Ultrasound technology

Background:

• Acoustic radiation force (ARF) imaging uses tissue displacement to measure mechanical properties.
• Tracking micrometer-level ARF-induced displacements in vivo is challenging.
• Bayesian estimation offers advantages over traditional methods like normalized cross-correlation (NCC).

Purpose of the Study:

• To develop and evaluate a novel Bayesian framework for enhanced displacement estimation in ARF imaging.
• To integrate a generalized Gaussian-Markov random field (GGMRF) prior with automated prior width selection.
• To assess the performance of this framework in tracking micro-displacements for elasticity imaging.

Main Methods:

• Developed a Bayesian framework incorporating a GGMRF prior and an automated prior width selection method.
• Evaluated the estimator's performance using simulations and in vivo cardiac radio-frequency ablation ARFI imaging data.
• Compared the proposed estimator against NCC, median-filtered NCC, and a previous Bayesian estimator.

Main Results:

• The proposed Bayesian estimator achieved up to a one order-of-magnitude reduction in mean-square error (MSE) at the automatically selected prior width.
• Lesion simulations demonstrated higher contrast-to-noise ratio but lower contrast compared to existing methods.
• In vivo results showed quantitative improvements in lesion contrast-to-noise ratio for ARFI imaging.

Conclusions:

• The developed Bayesian framework with GGMRF prior and automated width selection improves micro-displacement tracking accuracy in ARF imaging.
• The estimator offers enhanced lesion contrast-to-noise ratio in ARFI applications.
• This approach advances quantitative tissue property measurement in medical ultrasound.