A Parallel PDE-Constrained Optimization Framework for Biomedical Hyperthermia Treatment Planning

We present a PDE-constrained optimization algorithm which is designed for parallel scalability on distributed-memory architectures with thousands of cores. The method is based on a linesearch interior-point algorithm for large-scale continuous optimization, it is matrix-free in that it does not require the factorization of derivative matrices. Instead, it uses a new parallel and robust iterative linear solver on distributed-memory architectures. We will show almost linear parallel scalability results with 256 cores for the optimization problem, which is an emerging biomedical application and is related to antenna identi cation in hyperthermia cancer treatment planning. Additionally, we will discuss promising existing parallel matching algorithms which could further enhance the parallel scalability of general nonlinear optimization solvers.