Microwave Imaging Algorithms For Detection Of Breast Cancer In The Time Domain

Breast cancer is a sort of cancerous disease characterized as high incidence rate and high mortality, which threatens women’s physical and mental health. Thus, its early-stage detection is a critical way to reduce death rate.There, however, are some disadvantages in some case for conventional techniques such as X-ray mammogram, computerized tomography (CT), magnetic resonance imaging (MRI) and ultrasonic imaging, used for detection of breast cancer, and these disadvantages limit the mass screening of this disease.Microwave imaging (MWI), one of means used for nondestructive testing, is widely applicable. There is significant difference in electrical properties between normal breast tissues and malignant tumors at the microwave frequency range. This difference is the rationale of microwave imaging for breast tumors detection. Thus, it is a promising approach to provide a conventional or auxiliary detection means with low cost, little risk and easy mass screening program.Currently, three leading microwave imaging methods, microwave tomography imaging (MTI), confocal microwave imaging (CMI) and microwave-induced thermo-acoustic imaging (TAI), have been applied to detect early-stage breast tumors. And gratifying progress has been made at home and abroad, too.In this doctoral dissertation, the time-domain methodology is selected and the main achievement include: 1) Research on the propagation properties of microwave within breast; 2) Description of the frequency-dependent properties of breast tissues and tumors with the help of the single-pole Debye dispersive model; 3) Presenting of several one-dimensional (1-D), two-dimensional (2-D) and three-dimensional (3-D) microwave imaging approaches with high precision and rapid imaging to extract tumors knowledge and to detect breast cancer at early stage, effectively; 4) Using of the first-order Tikhonov’s regularization scheme to suppress the ill-posedness of the imaging problems and to cope with noise; and 5) Carrying out some numerical examples based on the finite-difference time-domain (FDTD) method and optimizing of adjustable parameters.The simulated results of aforementioned examples demonstrate that: 1) The proposed microwave imaging algorithms are robust; 2) The ultra-wideband (UWB) microwave imaging or microwave-induced thermo-acoustic imaging technique offers a high imaging contrast and high spatial resolution; and 3) Those microwave imaging techniques are feasible for early-stage breast cancer detection.These theoretical works provide basis for medical diagnosis and clinical application in the future.