Electromagnetic inversion for noninvasive specific absorption rate characterization

Loading...
Thumbnail Image
Date
2019
Authors
Phaneuf, Mario Lucien
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis investigates the use of the electromagnetic (EM) inversion framework for noninvasive specific absorption rate (SAR) characterization. Under this framework, the SAR characterization problem is formulated as an inverse problem where EM fields external to the SAR phantom are measured and then used to infer the internal SAR distribution. The advantages of the noninvasive approach include a potentially accelerated measurement, the option to use solid and inhomogeneous phantoms, and the ability to use existing near-field measurement hardware. An overview of contemporary SAR measurements and direct acceleration methods is presented to motivate the noninvasive approach. Following this, a classification of possible approaches to the noninvasive SAR characterization problem is presented and analyzed for the first time. From these approaches, two methods are chosen to be evaluated in this thesis. The first method is known as the phantom surface current method, which replaces the SAR phantom with equivalent surface currents which radiate in free space and can be used to obtain the internal SAR distribution. The noninvasive SAR problem is thus transformed into a near-field measurement of two independent antennas. A near-field antenna measurement algorithm is adapted for the SAR problem by including some novel numerical techniques. The performance is evaluated for 2D transverse magnetic (TM) problems. In addition, a fully vectorial 3D algorithm is developed and preliminarily investigated. The second method is proposed for the first time and is referred to as the simultaneous inversion (SI) method. This method is built upon the phantom surface current method by also considering the dependency between the two equivalent antennas and using this relationship as a regularization term. It is argued that this formulation is more resistant to measurement noise when compared to existing methods. The theoretical framework motivating this proposition is also presented. The noise resistance of the SI method and the theoretical framework are verified for 2D TM problems using synthetic measurement data perturbed by synthetic white noise.
Description
Keywords
Electromagnetic Inversion, Inverse Problems, Specific Absorption Rate, SAR, Noninvasive SAR
Citation