Calibration for hyperthermia treatment planning

Abstract

The current state of the art in electromagnetic hyperthermia requires a patient to be placed in a specific position and maintain this position during a treatment. As hyperthermia treatments may last an hour or more, this introduces an unpleasant constraint. Also, any deviations from the specified position by the patient may have a large impact on the effectiveness of the treatment. We investigate potentially interesting methods for calibrating the hyperthermic applicator as a first step to accommodating the potential displacement of the patient. Specifically, we examine the use of electromagnetic and power probes placed externally to the patient and investigate a technique by Fokkema and van den Berg [Fokkema and van den Berg(1996)] that may prove useful in a more advanced phase of calibration research. The probes mentioned are used to calculate the Green's function matrix relating their values to the applicator settings. This problem is then inverted to find the settings that will restore the original probe values (or a close approximation). The sensitivities analysis of Fokkema and van den Berg [Fokkema and van den Berg(1996)] uses Lorentz reciprocity and an assumption termed the Born approximation to determine whether (the in our case electromagnetic) sensitivities may be linearized with respect to a contrast - a perturbation from the original situation. We find that external probes equal in number to the number of antennas on the applicator insufficiently represent the fields inside the patient, especially in terms of tumor coverage and hotspots. The sensitivities analysis, however, proves promising on a toy problem and has potential to be expanded to a more complex configuration.

Date of Award	31 Aug 2015
Original language	English
Supervisor	Bas P. de Hon (Supervisor 1)