Experimental Closed-Loop Control of Breast Cancer in Mice

Abstract

Cancer therapy optimization is an issue that can be solved using the control engineering approach. An optimal therapy generation algorithm is presented and tested using a tractable mouse model of breast cancer. The optimized therapeutic protocol is calculated in a closed-loop manner at fixed time instants, twice in a week. The controller consists of a nonlinear model predictive controller which uses the state estimation of a moving horizon estimator. The estimator also computes parameter estimates of the prediction model such that the time varying nature of tumor evolution can be captured. Results show that remission can be induced in a 28-day interval using the algorithm.