Tensor network methods for Radiotherapy Dose Optimization

The very first successful application of tensor network methods (TNMs) to the solution of the dose optimization in Intensity Modulated Radiation Therapy (IMRT) is considered. This technique provides a method to modulate the local beam intensities, dividing the beam into smaller beamlets. This allows to reduce the radiation toxicity for healthy organs and deal with irregular and inhomogeneous tumors. Plan's goals are encoded as mathematical constraints into a cost function expressing the distance between the prescribed and delivered dose. Aim of the optimization is to minimize the function associating to each beamlet the optimal weight xj. In this thesis a classical quadratic cost function is mapped into an Insing-like Hamiltonian, where the beamlets weights are described by a system of long-range interacting qubits. The aim of the work is to solve the dose optimization problem using a binary tree tensor network (bTTN) to find the Hamiltonian's groundstate and to show the applicability of TNM to the IMRT dose optimization problem.