Adaptive Hybrid Optimization Algorithm with combination of search strategies for chemotherapy scheduling

Abstract

The primary challenge in oncology remains the design of chemotherapy schedules that effectively eradicate tumors without inducing lethal systemic toxicity or drug resistance. Existing metaheuristic optimization techniques often suffer from premature convergence and computation stagnation, failing to find global optima in the complex, high-dimensional search spaces of patient-specific protocols. To overcome these computational shortcomings, this study introduces the Adaptive Hybrid Optimization Algorithm (AHOA), a new hybrid optimization framework that transcends the limitations of static models by dynamically integrating five distinct metaheuristic strategies through an intelligent, state-triggered switching logic. The study employs a comprehensive bio-mathematical approach, incorporating multi-compartment pharmacokinetics, logistic tumor growth, and hematological toxicity models to simulate a 10-day treatment cycle. The objective function is formulated to minimize tumor volume while strictly enforcing clinical penalties for cumulative dosage and neutrophil suppression. Simulations performed in this study demonstrate that the adaptive algorithm achieves a superior tumor reduction of 10 . 9 mm 3 , maintaining 100% convergence reliability across 100 independent trials, a significant improvement over the erratic performance of standalone algorithms. Crucially, the algorithm maintained robust safety compliance, limiting cumulative doses to 4.88 units ( < 5 . 0 limit) and preserving neutrophil counts at 4 . 9 × 1 0 9 / L . These outcomes of this study suggest that the AHOA provides a reliable computational bridge for precision medicine; clinically, the results advocate for an early-intensification dosing strategy to exploit peak tumor growth kinetics. This framework offers a scalable, predictable tool for oncologists to transition from standardized regimens to tailored biological interventions.