Date of Graduation

12-2015

Document Type

Thesis

Degree Name

Master of Science in Computer Science (MS)

Degree Level

Graduate

Department

Computer Science & Computer Engineering

Advisor

John M. Gauch

Committee Member

Gordon Beavers

Second Committee Member

Christophe Bobda

Third Committee Member

Magda O. El-Shenawee

Abstract

Terahertz imaging technology has grown substantially in recent years. With new advances made in the generation, manipulation, and detection of terahertz radiation, terahertz imaging has found its application as inspection tools in a wide range of areas including biomedicine, materials science, and aerospace industries. In the field of medical imaging, terahertz (THz) has emerged as a new imaging modality due to some fundamental physical properties of the THz wave. THz imaging has been used to examine various biological structures, and THz tomography has been developed to probe the 3D structures of those samples.

THz pulsed imaging (TPI) is one of the THz imaging technologies that have been developed and used to inspect different types of biological structures. Specifically, it has been used to examine breast cancer tumors in hope of providing accurate, fast, and convenient diagnosis. Several works have been published in this research area that demonstrate the potency of THz imaging in differentiating tumor regions from healthy tissue regions. Most of the current research, however, focuses on the improvement of the instrumentation of THz imaging system or the study of the electrical and optical properties of the tissue samples being imaged. In this work, we seek to analyze the scanned image data and improve the visualization of them through image processing and segmentation techniques. We apply a variety of image enhancement and segmentation algorithms to our gathered data and evaluate how effective they are at visualizing the scanned samples and classifying the tissue regions. Our experimental results show that several image processing techniques can be utilized to help us visualizing the detailed structures of the scanned tissues and extract useful information which can be used for identifying tumor boundaries. The result of this work further demonstrates the possibility of adapting THz imaging system in real-world clinical settings that may facilitate the diagnosis of breast cancer.

Share

COinS