Spatially variant regularization of lateral displacement measurement using variance

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Ultrasound Elastography

Background:

• Accurate measurement of tissue displacement is crucial for shear modulus reconstruction in ultrasound elastography.
• Conventional methods often suffer from spatially varying accuracy, particularly in lateral displacement, impacting reconstruction stability.
• Existing spatially uniform regularization techniques do not fully address these accuracy variations.

Purpose of the Study:

• To confirm the effectiveness of a proposed spatially variant displacement component-dependent regularization method.
• To enhance the accuracy of two-dimensional (2D) lateral displacement measurements using ultrasonic methods.
• To improve the stability and accuracy of 2D shear modulus reconstruction.

Main Methods:

• The study proposes and validates a spatially variant lateral displacement regularization technique.
• This method is applied to previously developed ultrasonic 2D displacement vector measurement techniques: 2D cross-spectrum phase gradient method (CSPGM), 2D autocorrelation method (AM), and 2D Doppler method (DM).
• Effectiveness was verified through experiments on an agar phantom with a known shear modulus inclusion, using a 7.5-MHz linear array transducer.

Main Results:

• The proposed spatially variant regularization yielded more accurate lateral strain measurements compared to conventional uniform regularization.
• Improved detectability in lateral strain images was observed, indicated by higher contrast-to-noise ratios (CNRs) and signal-to-noise ratios (SNRs).
• More stable and accurate 2D shear modulus reconstructions were achieved, demonstrating the method's practical utility.

Conclusions:

• Spatially variant displacement component-dependent regularization significantly enhances lateral strain measurement and shear modulus reconstruction in ultrasound elastography.
• This improved accuracy and stability enable practical diagnostic and monitoring tools for soft tissue diseases and noninvasive therapies.
• Future application to elevational displacement could further enhance three-dimensional (3D) reconstruction stability.