This paper explores the impact of various decisions in estimating parameters from experimental tumour growth data on the outcome of the analysis using five mechanistic mathematical models (MMs) with increasing complexity. A framework is proposed to include censored data (measurements outside detection limits), demonstrating how its exclusion leads to biased estimations of initial tumour volume, carrying capacity, and model-predicted tumour volumes. The influence of prior choices on posterior distributions is highlighted, emphasizing the potential for misleading interpretations when only reporting most likely parameters and their 95% credible intervals.
Publisher
npj Systems Biology and Applications
Published On
Aug 14, 2024
Authors
Jamie Porthiyas, Daniel Nussey, Catherine A. A. Beauchemin, Donald C. Warren, Christian Quirouette, Kathleen P. Wilkie
Tags
tumor growth
parameter estimation
censored data
mechanistic mathematical models
analysis bias
Related Publications
Explore these studies to deepen your understanding of the subject.
