Predictive food microbiology: A probabilistic approach Original Research Article

An existing mathematical model for microbial growth and inactivation in dynamically changing temperature conditions has been modified to explain the experimentally observed variability between replicate growth curves. The model incorporates a random variable parameter representing the random physiological state of the inoculum. The model is validated using experimental growth data which were obtained in time-varying temperature conditions. It is shown by Monte Carlo analyses that broad confidence bands are associated with the model predictions.