MICROWAVE INVERSE RECONSTRUCTION METHODS FOR IMAGING

Open Access
- Author:
- Zhou, Huiyuan
- Graduate Program:
- Electrical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 20, 2018
- Committee Members:
- Ram Mohan Narayanan, Dissertation Advisor/Co-Advisor
Ram Mohan Narayanan, Committee Chair/Co-Chair
James Kenneth Breakall, Committee Member
Timothy Joseph Kane, Committee Member
Michael T Lanagan, Outside Member - Keywords:
- Microwave Imaging
Dual Mesh
Distmesh
Levenberg Marquardt
Iteratve Method with Adaptive Threshold
Non-decimated Wavelet Transformation - Abstract:
- In this thesis, various microwave inverse scattering methods are developed, which are able to be applied to biomedical applications, such as localization of capsule endoscope, breast tissue imaging, or brain stroke detection. The configuration applied in this thesis addresses external transmit/receive antennas which are placed around the object to be imaged. During operation, a single transmitter emits the microwave signal at each instant and all the other receivers collect the signal. As the result, the measured data from the system is in the form of a matrix, which is used for the reconstruction process. Based on the object to be imaged, the reconstruction image area is divided using a dual mesh grid, which is able to provide a higher resolution reconstruction result in the region of interest. This dual mesh method makes it possible to achieve a good balance between high accuracy and acceptable computational complexity. The reconstruction algorithm includes two parts: the forward problem and the inverse problem. The Method of Moments (MoM) is selected as the forward solver cooperating with two mesh types, the rectangular mesh and the triangular mesh. In addition, the incident field is set as TM polarization. For inverse method, conventional nonlinear optimization method, Levenberg Marquardt, and innovative NDWT-IMATCS method are introduced. Simulation results are presented to verify the performance of the algorithms, which are applied on both the synthetic simulated data and practical experimental data. Furthermore, the performance comparison and analysis of the algorithms are presented. Finally, the CoSaMP algorithm is proposed as future work.