Quantitative elasticity imaging with ultrasound

Abstract

For the past 10 or so years, various researchers have proposed different ways to measure strain distributions in vivo. The measured strains result from various sources including external palpation, low frequency vibration, or internal motion. By examining the relative strains in adjacent tissues, it is thought that one can infer the relative distributions of tissue stiffness. In this presentation, we discuss the process of creating a quantitative stiffness image from a given measurement of tissue strain. We show that the strain image by itself is insufficient information to infer the elastic stiffness. By examining the well-posedness of the inverse problem, we determine what information is needed to supplement the strain image in order to quantitatively infer the elastic stiffness distribution. Methods for obtaining the required information in vivo are currently being developed. We show examples of incorrect and misleading stiffness reconstructions in the absence of the required data. [Work supported by BU, NIH, Fulbright Foundation, ICR, and NSF via CenSSIS ERC.]