Though clinicians can now collect detailed information about a variety of tumor characteristics as a tumor evolves, it remains difficult to predict the efficacy of a given treatment prior to administration. Additionally, the process of data collection may be invasive and expensive. Thus, the creation of a framework for predicting patient response to treatment using only information collected prior to the start of treatment could be invaluable. In this study, we employ ordinary differential equation models for tumor growth and utilize synthetic data from a cellular automaton model for calibration. We investigate which parameters have the most influence upon treatment efficacy by comparing parameter distributions associated with treatment outcomes. Additionally, we develop a framework for estimating the probability of observing complete tumor remission following a simulated radiotherapy regimen based only on a patient’s non-treatment parameters, so that treatment efficacy could be predicted prior to administration.
Huang, Yutian and Lewis, Allison L.
"Predicting Tumor Response to Radiotherapy Based on Estimation of Non-Treatment Parameters,"
Spora: A Journal of Biomathematics: Vol. 7, 25–35.
Available at: https://ir.library.illinoisstate.edu/spora/vol7/iss1/5