Operator splitting for adaptive radiation therapy with nonlinear health dynamics

We present an optimization-based approach to radiation treatment planning over time. Our formulates as optimal control problem with nonlinear patient health dynamics derived from the standard linear-quadratic cell survival model. As formulation is nonconvex, we propose a method for obtaining approximate solution by solving sequence of convex optimization problems. This fast, efficient, and robust model error, adapting readily changes in patient's between sessions. Moreover, show that it can be combined operator splitting ADMM produce algorithm highly scalable handle large clinical cases. introduce open-source Python implementation our algorithm, AdaRad, demonstrate its performance on several examples.