Date of Completion

10-25-2017

Embargo Period

12-31-2017

Keywords

Medical imaging, Biomedical Optics, Breast Cancer, Near Infrared Imaging, Ultrasound, Tissue Optics

Major Advisor

Quing Zhu

Co-Major Advisor

Rajeev Bansal

Associate Advisor

Monty A. Escabi

Associate Advisor

Guoan Zheng

Associate Advisor

Patrick Kumavor

Field of Study

Biomedical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

According to the World Health Organization, breast cancer is the most common cancer among women worldwide, claiming lives of hundreds of thousands of women each year. Mammography, ultrasound (US) and magnetic resonance imaging (MRI) are widely used to detect and diagnose breast lesions and each of them have some disadvantages which motivates researchers to find a new and high sensitive imaging method. Diffuse optical tomography (DOT) is a noninvasive functional imaging modality that utilizes near-infrared (NIR) light to probe tissue optical properties. Minimal light absorption in the NIR spectrum allows for several centimeters of light penetration in soft tissue. NIR is sensitive to hemoglobin which is directly related to tumor angiogenesis. If multiple wavelengths are used, it can probe tumor oxygen saturation. However, DOT has drawbacks on location uncertainty and low light quantification accuracy due to tissue scattering of NIR light. Ultrasound (US)-guided Diffuse Optical Tomography has overcome these problems.

The overall objective of my research is to move US-guided DOT one step closer from a research imaging modality to a commercial prototype for future wide use in clinics. To achieve this goal, I have overcome several challenges in both imaging hardware and software. In this Dissertation, I will discuss the progress we have made on development of two generations of US-guided DOT systems along with the software improvements. The developed systems have been improved in terms of robustness, stability, clinical safety and standards and user friendliness. In terms of software and algorithm development, automated methods for data selection, outlier removal and reference and perturbation filtering are presented. These methods have improved the robustness of the technique and increased speed of image formation by eliminating the need for an expert to perform data preprocessing and data selection. Phantom and clinical experimental results will be demonstrated to evaluate the performance of the developed systems and algorithms.

COinS