Quantitative Ultrasound and the Effects of Acoustic Nonlinearity

Abstract

El desarrollo de nuevas técnicas de imagen utilizando señales de ultrasonido que tienen diferentes fuentes de contraste de imagen es médicamente significativo. Por ejemplo, aquellas basadas en el mapeo de parámetros de ultrasonido cuantitativo (QUS). Se sabe además que diferentes tejidos tienen diferentes coeficientes de no linealidad (B/A), que pueden distorsionar en menor o mayor grado a los estimados de QUS. En la primera parte de la tesis, se determinó las condiciones que pueden disminuir la distorsión no lineal de los estimados de QUS provocada por diferenies niveles de presión acústica y se exploró un método de calibración in situ que minimizaría tal distorsión no lineal. En la segunda parte del trabajo, se presentaron dos métodos para calcular el B/A basados en observaciones relacionadas con estimados de QUS. El primer método fue un enfoque heurístico basado en la estimación del exceso del coeficiente de atenuación. El segundo método para la estimación del B/A se realizó en el dominio del tiempo utilizando el principio de conservación de la energía y un cálculo implícito de la señal del segundo armónico. Usando este último método se logró construir imágenes paramétricas del promedio acumulado de B/A en medios heterogéneos.

Diagnostic ultrasound is the most heavily utilized imaging modality in medicine worldwide second only to digital X-ray. Conventional B-mode imaging relies on small impedance differences (often <5%) between tissues to provide image contrast. This perceived contrast is further reduced because B-mode images are replete with speckle. The development of novel imaging techniques using ultrasound signals that have different sources of image contrast and may not be affected by speckle is medically significant; for example those based on mapping quantitative ultrasound (QUS) parameters. However, different types of tissues or tissues under different disease states are known to have different coefficients of nonlinearity (B/A). For example, fatty tissues have a B/A of 11, liver has a B/A of 6.6, and water has a B/A of 5. At low pressures it can be assumed that QUS parameters are not changed by acoustic nonlinear distortion, whereas at higher pressures the nonlinear distortion transfers energy from the fundamental frequency of the ultrasound wave into higher harmonics. These distortions can affect the bias and variance of spectralbased QUS estimates, such as the backscatter coefficient and attenuation coefficient. In the first part of this dissertation, we aim to determine conditions that can decrease acoustic nonlinear effects. We explored an in situ calibration approach which would minimize nonlinear distortion on QUS estimates. In the second part of this dissertation we present two methods to calculate the B/A based on observations related to QUS estimation. The first method was an heuristic approach based on estimating the excess attenuation coefficient and required two well-characterized reference phantoms. The second method for B/A estimation needed only one reference phantom and was performed mainly in the time-domain using the conservation of energy principle. Using the later method we can construct parametric images to map the cumulative average B/A versus depth.