Related Experiment Video
Updated: Sep 11, 2025
10:33
Expedited Radiation Biodosimetry by Automated Dicentric Chromosome Identification ADCI and Dose Estimation
Published on: September 4, 2017
15.9K
A Scoping Review on Calibration Methods for Cancer Simulation Models
Yichi Zhang1, Nicole Lipa1, Oguzhan Alagoz1
1Department of Industrial and Systems Engineering, University of Wisconsin-Madison, Madison, WI, USA.
Summary
This review of cancer simulation models found that while calibration targets are commonly reported, parameter search algorithms like random search are predominantly used. Machine learning algorithms remain underutilized in this field.
Area of Science:
- * Computational epidemiology
- * Health informatics
- * Cancer modeling
Background:
- * Calibration is essential for simulation models, especially in cancer research where natural history data is scarce.
- * Accurate calibration ensures model outcomes align with observed data, crucial for reliable predictions.
Purpose of the Study:
- * To conduct a scoping review of calibration methods in cancer simulation models with a natural history component.
- * To identify and analyze the calibration approaches, targets, and algorithms used in the literature.
Main Methods:
- * A comprehensive scoping review of studies from 1980 to August 2024.
- * Keyword searches in PubMed and Web of Science for cancer simulation models incorporating calibration.
- * Inclusion criteria focused on models with natural history components and parameter estimation via calibration.
Main Results:
- * 117 studies were included; most specified calibration targets (incidence, mortality, prevalence) and parameter search algorithms.
- * Goodness-of-fit metrics, acceptance criteria, and stopping rules were less frequently reported.
- * Random search was the most common parameter search algorithm, followed by Bayesian and Nelder-Mead methods.
Conclusions:
- * Calibration practices in cancer simulation models are established, with common targets and algorithms.
- * Machine learning algorithms are underutilized despite their potential for improving calibration efficiency.
- * Further research is recommended to compare the performance of various parameter search algorithms.
