Optimization of Multi-Stage Neuroblastoma Therapy Based on Differential Transformations

A numerical-analytical method for optimizing multi-stage high-risk neuroblastoma therapy is proposed, based on the differential transform method and a terminal control framework. The mathematical model of the “tumor-neuropil-immunity-pharmacokinetics” system is formulated in spectral form, enabling optimal control synthesis without numerical integration of the system of differential equations. The proposed approach yields a compact recurrent representation of the dynamics, an analytical form of the optimal control law, and a closed-loop control algorithm robust to parameter perturbations and individual patient responses. It is shown that the method provides significant computational acceleration compared to gradient-based optimization, making it suitable for near-real-time clinical decision support systems.