## Abstract

The methods used for biomedical imaging include X-ray, magnetic resonance imaging (MRI), ultrasound, and most recently, microwave imaging each with its own advantages and disadvantages for specific application. The frequency dependent nature of the Electromagnetic (EM) waves makes it possible for microwave imaging to achieve a tissue differentiation comparable to that obtained by X-ray and would be superior to ultra sound and MRI.

This work addresses one aspect of an ongoing research effort at Rochester Institute of Technology which is a new methodology for detecting diseased tissues by a non-invasive procedure. The methodology is based on the extraction of the frequencydependent electrical properties of tissue that can differentiate between healthy and diseased states using ultra-wideband (UWB) pulses that are allowed to impinge on multiple layers of biological tissue. The reflected/scattered signal is analyzed to obtain frequency dependent permittivity and conductivity that correspond to each tissue layer. For system implementation, a planar flexible antenna array is utilized that would be wrapped like a cuff around a specific part of the human body.

The present work addresses the development of a calibration procedure to locate each antenna element with respect to the shape of the body surface. An algorithm for determining the shape of the surface has been developed in this thesis. The methodology involves placing a flat N x N planar antenna array above the tissue surface. The antenna elements are excited one at a time and the rest of the elements are used to receive signals scattered from the body surface. From the transmitted and received signals, the distance, r, from each antenna to the surface can be determined and in this work it is assumed to be known. Corresponding to this distance the sphere of radius r_{ij }is drawn from the (i,j)^{th} antenna element. Assuming that the surface is convex, this sphere will be tangent to the body surface at the point with coordinates (x_{ij}, y_{ij}, z_{ij}). Due to the fact that the antennas are placed relatively close to each other, it can be assumed that the parameters of the neighboring tangent planes are the same. Using this information together with the distances r_{ij}, r_{(i+1)j} and r_{(i,j+1)}, the necessary equations have been set up and solved for the coordinates (x_{ij}, y_{ij}, z_{ij}). This process is repeated for N xN sample points on the surface which are then used to determine the surface by an interpolation algorithm. The algorithm is implemented in MATLAB and its accuracy has been assessed for different kinds of surfaces and different sized planar antenna array. The agreement between the estimated and the actual surfaces is very good for smooth convex surfaces. For surfaces with more curvature, the mean square error is higher. The performance of the algorithm with respect to the measurement noise has also been analyzed.

After the planar antenna array has been calibrated with respect to the surface, the frequency response of the paths from the antenna to each tissue layer including the surface can be obtained. This information is then used for an accurate tissue differentiation.

## Library of Congress Subject Headings

Imaging systems in medicine; Tissues--Imaging; Tissues--Optical properties; Microwave imaging

## Publication Date

2006

## Document Type

Thesis

## Student Type

Graduate

## Degree Name

Electrical Engineering (MS)

## Department, Program, or Center

Electrical Engineering (KGCOE)

## Advisor

Sohail Dianat

## Advisor/Committee Member

Jayanti Venkataraman

## Advisor/Committee Member

Eric Peskin

## Recommended Citation

Teerakapibal, Surat, "Tissue Surface Identification for Microwave Imaging System Calibration" (2006). Thesis. Rochester Institute of Technology. Accessed from

https://repository.rit.edu/theses/8019

## Campus

RIT – Main Campus

## Comments

Physical copy available from RIT's Wallace Library at R857.O6 T44 2006