Inverse extremal problem for an anti-tumor therapy model

Andrey Kovtanyuk,Christina Kuttler,Kristina Koshel,Alexander Chebotarev
2024-07-11
Abstract:An optimal control problem for a model of tumor growth is studied. In a given subdomain, it is required to minimize the density of tumor cells, while the drug concentration in tissue is limited by given minimal and maximal values. Based on derived estimates of the solution of the controlled system, the solvability of the control problem is proved. The problem is reduced to an optimal control problem with a penalty. An algorithm for solving the optimal control problem with a penalty is constructed and implemented. The efficiency of the algorithm is illustrated by a numerical example.
Optimization and Control
What problem does this paper attempt to address?
This paper mainly discusses an inverse optimization problem for a tumor drug treatment model. Existing mathematical models are usually based on ordinary differential equations (ODEs) or partial differential equations (PDEs) to describe the evolution of tumors, but they are complex and involve multiple state variables. The researchers considered a simplified model with two state variables: the normalized density of tumor cells and the normalized concentration of the drug. The paper proposes an optimal control problem, aiming to minimize the density of tumor cells within a given subdomain while restricting the drug concentration to specific minimum and maximum values. The authors prove the existence of the solution to the optimal control problem and propose an algorithm based on penalty functions to solve it. The effectiveness of the algorithm is demonstrated through numerical examples. In addition, the paper discusses the uniqueness and asymptotic properties of the solution to the problem, and provides algorithmic steps for numerical solution, including the iterative process and convergence. Finally, the effectiveness of the model and algorithm is validated through a computational example, simulating the variation of tumor cell density over time. Future research will extend to include nonlinear reaction terms and models describing multiple types of tumor cells.