Uncertainty in the sizing of breast cancer masses is a major issue in breast screening programs, as there is a tendency to severely underestimate the sizing of malignant masses, especially with ultrasound imaging as part of the standard triple assessment. Due to this issue about 20% of all surgically treated women have to undergo a second resection, therefore the aim of this thesis is to address this issue by developing novel image analysis methods. Ultrasound elasticity imaging has been proven to have a better ability to differentiate soft tissues compared to standard B-mode. Thus a novel segmentation algorithm is presented, employing elasticity imaging to improve the sizing of malignant breast masses in ultrasound. The main contributions of this work are the introduction of a novel filtering technique to significantly improve the quality of the B-mode image, the development of a segmentation algorithm and their application to an ongoing clinical trial. Due to the limitations of the employed ultrasound device, the development of a method to improve the contrast and signal to noise ratio of B-mode images was required. Thus, an autoregressive model based filter on the radio-frequency signal is presented which is able to reduce the misclassification error on a phantom by up to 90% compared to the employed device, achieving similar results to a state-of-the art ultrasound system. By combining the output of this filter with elasticity data into a region based segmentation framework, a computationally highly efficient segmentation algorithm using Graph-cuts is presented. This method is shown to successfully and reliably segment objects on which previous highly cited methods have failed. Employing this method on 18 cases from a clinical trial, it is shown that the mean absolute error is reduced by 2 mm, and the bias of the B-Mode sizing to underestimate the size was overcome. Furthermore, the ability to detect widespread DCIS is demonstrated.

Posterior tibial tendon dysfunction (PTTD) is a common degenerative condition leading to a severe impairment of gait. There is currently no effective method to determine whether a patient with advanced PTTD would benefit from several months of bracing and physical therapy or ultimately require surgery. Tendon degeneration is closely associated with irreversible degradation of its collagen structure, leading to changes to its mechanical properties. If these properties could be monitored in vivo, it could be used to quantify the severity of tendonosis and help determine the appropriate treatment. Ultrasound elasticity imaging (UEI) is a real-time, noninvasive technique to objectively measure mechanical properties in soft tissue. It consists of acquiring a sequence of ultrasound frames and applying speckle tracking to estimate displacement and strain at each pixel. The goals of my dissertation were to 1) use acoustic simulations to investigate the performance of UEI during tendon deformation with different geometries; 2) develop and validate UEI as a potentially noninvasive technique for quantifying tendon mechanical properties in human cadaver experiments; 3) design a platform for UEI to measure mechanical properties of the PTT in vivo and determine whether there are detectable and quantifiable differences between healthy and diseased tendons. First, ultrasound simulations of tendon deformation were performed using an acoustic modeling program. The effects of different tendon geometries (cylinder and curved cylinder) on the performance of UEI were investigated. Modeling results indicated that UEI accurately estimated the strain in the cylinder geometry, but underestimated in the curved cylinder. The simulation also predicted that the out-of-the-plane motion of the PTT would cause a non-uniform strain pattern within incompressible homogeneous isotropic material. However, to average within a small region of interest determined by principal component analysis (PCA) would improve the estimation. Next, UEI was performed on five human cadaver feet mounted in a materials testing system (MTS) while the PTT was attached to a force actuator. A portable ultrasound scanner collected 2D data during loading cycles. Young's modulus was calculated from the strain, loading force and cross sectional area of the PTT. Average Young's modulus for the five tendons was (0.45±0.16GPa) using UEI. This was consistent with simultaneous measurements made by the MTS across the whole tendon (0.52±0.18GPa). We also calculated the scaling factor (0.12±0.01) between the load on the PTT and the inversion force at the forefoot, a measurable quantity in vivo. This study suggests that UEI could be a reliable in vivo technique for estimating the mechanical properties of the human PTT. Finally, we built a custom ankle inversion platform for in vivo imaging of human subjects (eight healthy volunteers and nine advanced PTTD patients). We found non-linear elastic properties of the PTTD, which could be quantified by the slope between the elastic modulus (E) and the inversion force (F). This slope (ΔE/ΔF), or Non-linear Elasticity Parameter (NEP), was significantly different for the two groups: 0.16±0.20 MPa/N for healthy tendons and 0.45±0.43 MPa/N for PTTD tendons. A receiver operating characteristic (ROC) curve revealed an area under the curve (AUC) of 0.83±0.07, which indicated that the classifier system is valid. In summary, the acoustic modeling, cadaveric studies, and in vivo experiments together demonstrated that UEI accurately quantifies tendon mechanical properties. As a valuable clinical tool, UEI also has the potential to help guide treatment decisions for advanced PTTD and other tendinopathies.

<p>The rupture of arterial plaques is the most common cause of ischemic complications including stroke, the fourth leading cause of death and number one cause of long term disability in the United States. Unfortunately, because conventional diagnostic tools fail to identify plaques that confer the highest risk, often a disabling stroke and/or sudden death is the first sign of disease. A diagnostic method capable of characterizing plaque vulnerability would likely enhance the predictive ability and ultimately the treatment of stroke before the onset of clinical events.</p><p>This dissertation evaluates the hypothesis that Acoustic Radiation Force Impulse (ARFI) imaging can noninvasively identify lipid regions, that have been shown to increase a plaque's propensity to rupture, within carotid artery plaques <italic>in vivo</italic>. The work detailed herein describes development efforts and results from simulations and experiments that were performed to evaluate this hypothesis.</p><p>To first demonstrate feasibility and evaluate potential safety concerns, finite-element method simulations are used to model the response of carotid artery plaques to an acoustic radiation force excitation. Lipid pool visualization is shown to vary as a function of lipid pool geometry and stiffness. A comparison of the resulting Von Mises stresses indicates that stresses induced by an ARFI excitation are three orders of magnitude lower than those induced by blood pressure. This thesis also presents the development of a novel pulse inversion harmonic tracking method to reduce clutter-imposed errors in ultrasound-based tissue displacement estimates. This method is validated in phantoms and was found to reduce bias and jitter displacement errors for a marked improvement in image quality <italic>in vivo</italic>. Lastly, this dissertation presents results from a preliminary <italic>in vivo</italic> study that compares ARFI imaging derived plaque stiffness with spatially registered composition determined by a Magnetic Resonance Imaging (MRI) gold standard in human carotid artery plaques. It is shown in this capstone experiment that lipid filled regions in MRI correspond to areas of increased displacement in ARFI imaging while calcium and loose matrix components in MRI correspond to uniformly low displacements in ARFI imaging.</p><p>This dissertation provides evidence to support that ARFI imaging may provide important prognostic and diagnostic information regarding stroke risk via measurements of plaque stiffness. More generally, the results have important implications for all acoustic radiation force based imaging methods used clinically.</p> / Dissertation