Real Time Elastography and Parametric US Imaging System

University Of Tennessee Health Sci Ctr, Memphis TN

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): According to the Guide to Clinical Preventive Services, the common form of screening for thyroid abnormalities, neck palpation, lacks the sensitivity (33%) for detecting thyroid cancer, and ultrasound has too great a false positive rate to avoid invasive tests such as biopsy, to rule out cancer. It is the aim of this work to provide sensitive ultrasound tests to non invasively diagnose cancer in the thyroid earlier and more accurately than current imaging techniques. Elastography, which images the stiffness properties of tissues, and parametric backscatter imaging that displays the frequency dependence of ultrasound reflections offer novel signal processing approaches that provide new information about tissues. These advanced ultrasound techniques will be incorporated into a new, software-controlled research ultrasound platform. Phase I efforts will tune the platform so that it provides data and software control needed to fully develop and evaluate these novel imaging methods. Further work will evaluate algorithms for producing elastograms using raw radio frequency echo data available on the research scanner. It also will develop acoustic parametric images that are based on the frequency dependence of scattering and on attenuation, offering modes that will compliment conventional amplitude mode processing. A software development kit (SDK) that accesses fast digital signal processing (DSP) boards on the scanner will also be developed. Quantitative images of thyroids will be derived to demonstrate the new modes. Phase II will implement elastograms in real time and provide other parametric images during image freeze. Real time elastography will be achieved through careful evaluation of different algorithms for use on the DSP circuit. Algorithms will be assessed for the quality of elastograms produced and the computational overhead incurred in their generation. Phantom and preclinical tests will be done to select the most effective for thyroid imaging. Parametric images based on attenuation and the frequency dependence of scattering, characteristic of the scatterer size will be developed. Novel features will be to incorporate reference phantom data for calibration directly into the scanner memory to facilitate rapid processing, and to apply compound scanning for reducing statistical fluctuations in quantitative images. The system will be evaluated by testing in patients with thyroid nodules and in patients with proven breast masses.