? The current practice of elastography, images only the axial component of the strain tensor, while the lateral and elevational strain tensor components are basically disregarded. In the absence of lateral and elevational components, other important elastic parameters such as shear strains and Poisson's are not imaged. Moreover, efforts at reconstruction of the Young's modulus from strain data would be significantly improved if all the normal and shear strain tensors are available. In this research, we propose a novel method of estimating displacements and strains in axial, lateral and elevational directions in ultrasound elastography imaging. The method uses tissue displacement components estimated along multiple angular ultrasound beam directions acquired with linear array transducers. Strain elastograms of normal and shear tensors are then calculated from the axial and lateral displacement information in addition to Poisson's ratio information. ? ? The research plan develops and optimizes this method for a state-of-the-art scanner using simulations to evaluate tradeoffs between acquisition rate, angles and data accuracy. Implementation of the techniques will be on the Siemens Antares scanner, capitalizing on a newly developed ultrasound research interface to facilitate angular beam data acquisition and control. Tests using elastography phantoms will yield performance benchmarks. Two new phantom types will also be developed to evaluate the ability to derive the fundamental information expected in these images. Finally, excised tissue samples will be imaged to evaluate the new elasticity information obtained. ? ? Clinical applications where the methodology will be directed are in differentiating malignant from benign masses in the breast and thyroid. In both sites lateral slippage is believed to occur in many benign masses when the tissue is compressed, while slippage appears to be absent with malignant masses. The proposed lateral and shear strain images are expected to be particularly sensitive to such phenomena. ? ?
Showing the most recent 10 out of 22 publications
