Efficient algorithm for computing optimal control of antennas in hyperthermia (Q1312220)

The authors present a model of energy distributions inside the tumour and inside the normal tissues for a finite number of microwave antennas used to heat up a cancer region in hyperthermia. The model is based on given data of a patient and is used to find phases and amplitudes which maximize the energy inside the tumour under the constraints imposed on the energy-deposition to the healthy tissues. To solve this problem the authors propose to transform it into a generalized eigenvalue problem for the linear mappings which are represented by two positive semidefinite integral operators \(A\) and \(B\). More precisely, the largest eigenvalue of \(B^{-1} A\) is determined by the proportion between the maximum supply of energy to the tumour and maximum strain for healthy tissue of the patient. The components of any eigenvector of the largest eigenvalue represent optimal control of antennas for the optimization problem. Assuming the infinite extension of antennas and media in \(z\)-direction the model is simplified to an uncoupled system of integral equations where a scalar equation is used for the electric field. Moreover, the authors present numerical experiments based on a hyperthermia system BSD-2000 used in the Rudolf Virchow Klinik in Berlin.