A Nonconvex Optimization Approach to IMRT Planning with Dose–Volume Constraints

Fluence map optimization for intensity-modulated radiation therapy planning can be formulated as a large-scale inverse problem with competing objectives and constraints associated the tumors organs-at-risk. Unfortunately, clinically relevant dose-volume are nonconvex, so standard algorithms convex problems cannot directly applied. While prior work focused on approximations these constraints, we propose novel relaxation approach to handle nonconvex constraints. We develop efficient, provably convergent based partial minimization, show how adapt them maximum-dose infeasible problems. demonstrate our using CORT dataset, that it is easily adaptable treatment multiple