Combined Stochastic Models for Cancer Patient Trajectories

Explore a computational approach for integrating diverse clinical data modalities in cancer patient analysis through a 24-minute conference talk at JuliaCon 2024. Delve into the development of a combined stochastic model that describes cancer patients' disease progression, incorporating continuous tumor growth, metastasis spread, and survival status. Learn how Bayesian inference is applied to estimate model parameters, and examine various models differing in their description of primary tumor growth. Discover the use of the SciML ecosystem in Julia for simulating continuous processes, coupled with a custom-implemented algorithm for jump processes. Gain insights into likelihood-based Bayesian inference frameworks, including MCMC sampling and particle filter algorithms. Understand the evaluation of model performance through simulation studies and the advantages of implementing this approach in Julia. Consider future extensions of this work, including the study of therapy effects and incorporation of mixed effects for inter-individual variability.