Files in this item

FilesDescriptionFormat

application/pdf

application/pdfLUCHIES-DISSERTATION-2016.pdf (7MB)
(no description provided)PDF

Description

Title:Using two-dimensional impedance maps to study acoustic properties of tissue microstructure
Author(s):Luchies, Adam C.
Director of Research:Oelze, Michael L
Doctoral Committee Chair(s):Oelze, Michael L
Doctoral Committee Member(s):O'Brien, William D; Insana, Michael F; Do, Minh
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Quantitative ultrasound
Backscatter coefficients
Impedance maps
Acoustic scattering
Abstract:Quantitative ultrasound (QUS) imaging represents a set of techniques for estimating acoustic properties of tissue microstructure that have confirmed potential for identifying disease and monitoring therapy. In one approach to QUS, the backscatter coefficient (BSC) from a tissue is utilized to quantify and classify tissue state. The BSC is a fundamental property of a tissue based on the frequency power spectrum estimated from the RF signals corresponding to ultrasonic backscatter. From the BSC, parametric models can be constructed to relate the frequency-dependent BSC to geometrical properties of the underlying tissue. However, most of these parametric models are based on analytic expressions (e.g., Gaussian function) and not on actual tissue morphology. The three-dimensional impedance map (3DZM) is a computational tool to create tissue specific form factor models directly from tissue histology. 3DZMs are constructed from a series of adjacent histological tissue slides that have been stained to emphasize acoustic impedance structures. The power spectrum of a 3DZM can be related to the BSC. Therefore, ZMs can be used to create tissue specific form factors. However, the process of constructing a 3DZM is expensive in terms of slide preparation time, computational time, and financial cost. In addition, there are multiple opportunities for large distortions to be introduced when constructing 3DZMs. A method based on analyzing two-dimensional impedance maps (2DZMs) would avoid many of the shortcomings of the 3DZM method. The proposed 2DZM method exploits the properties of isotropic media to estimate the correlation coefficient from slices before estimating the 3D volume power spectrum. Simulations were used to verify that 2DZMs could be used to estimate correlation coefficients and 3D power spectra having low error. The studied media had known correlation coefficients and power spectra so it was possible to verify that estimation of the correlation coefficient and power spectrum was possible. First, collections of sparse scatterers (e.g., spheres and ellipsoids) were studied. These studies indicated that correlation coefficients with RMSE less than 1% resulted when the 2DZMs contained 15 object cross sections. Second, media having a spherical Gaussian correlation coefficient and power spectrum were studied. This study indicated that correlation coefficients with RMSE less than 3% and power spectra with RMSE less than 11% resulted when using a single 2DZM having a size that was 50 times the scatterer size. Third, collections of dense spheres were studied. This study indicated that correlation coefficients with RMSE 1.5% resulted when using a single 2DZM having a size that was 25 times the scatterer size. Power spectra with RMSE 25% resulted when using 20 2DZMs having size that was 25 times the scatterer size. ZMs created from healthy rabbit livers were studied. An analysis of bias was carried out to determine the smallest size 2DZM that could be used without biasing the correlation coefficient and power spectral estimates. The results of this study indicated that correlation coefficients with RMSE 0.9% and power spectra with RMSE 1.4% resulted when using 2DZMs with side length 150 microns. An analysis of variance was carried out to determine the number of 2DZMs that needed to be used to reduce variance in the correlation coefficient and power spectral estimates. The results of this study indicated that correlation coefficients with RMSE 0.9% and power spectra with RMSE 1.4% resulted when estimating the correlation coefficient using 10 2DZMs. The 2DZM approach was tested on simulated media having known correlation coefficients and power spectra. The simulation results demonstrated that the 2DZM method was able to capture information about the size, shape, and 3D spatial locations of the scatterers. The rabbit liver results demonstrated the 2DZM method working with actual histology. These findings demonstrate that 2DZMs can be used to model ultrasonic scattering.
Issue Date:2016-09-01
Type:Thesis
URI:http://hdl.handle.net/2142/95272
Rights Information:Copyright 2016 Adam Luchies
Date Available in IDEALS:2017-03-01
Date Deposited:2016-12


This item appears in the following Collection(s)

Item Statistics