With the exponential growth of computer power, cancer screening modelers are constructing increasingly complex models that mimic real life clinical trials.  In a cancer screening model, parameters that govern critical but unobservable events (e.g., metastasis) are often estimated by calibrating the model to relevant data (e.g., from tumor registries and clinical trials).  As simulation models become more complex, calibrating them becomes both more important and more difficult.  In the field of engineering, many techniques have been cultivated to improve the speed and the quality of calibration.  However, cancer screening modelers utilize few of these techniques.  The goal of this project is to transfer and extend engineering simulation techniques to the field of cancer screening simulation modeling.

This project will apply engineering optimization techniques to calibrate numerous cancer screening models at the ITA.  We will integrate and automate these techniques to search the natural history parameter space and simultaneously calibrate to multiple endpoints.  We will measure, analyze, and compare the performance of these techniques.

This research will advance our understanding and methodologies of constructing and calibrating cancer screening models and will create streamlined procedures for model development.