Sensitivity and identifiability analysis of a third-order tumor growth model

Abstract

The growing cancer cases attract more and more scientific research and introductions of new models, applied control algorithms and methods. The models are fundamental in the area of computer generated low-dose metronomic (LDM) chemotherapy, when the administration of the drug is ought to be optimized. Generally the in-silico tests and investigations are based on a model, which is hypothesized to describe the given process reliably and accurately. The analysis of the models and its parameters is crucial for therapy generation. We performed an analysis of a third-order tumor growth model based on sensitivity analysis and identifiability tests. The results show that a subset of parameters can be fixed as population values and the rest of the parameter sets results in an identifiable system with minor loss of accuracy.