(626v) An Optimization Algorithm for Estimating Microbial Survival Parameters

The microbial survival curve can be described by an explicit linear or nonlinear equation of time when microorganisms are exposed to a lethal agent (e.g., temperature and high pressure) at constant intensity. Each model parameter involved in the equation is a function of the agent intensity, where parameters used to describe the model parameters are called microbial survival parameters. If the microbial survival parameters are known, then survival curves under changing agent intensity can be predicted. Therefore, the objective of this study was to develop an algorithm based on the steepest decent optimization method to estimate microbial survival parameters from survival data under known agent profiles, i.e., agent intensity versus time. The developed algorithm was validated by using the published thermal inactivation data of Salmonella. The data showed that Salmonella's survival curves followed the Weibull model, a popular nonlinear model. Using the data, Salmonella's survival parameters were identified from final survival ratios of three survival curves, single and multiple whole survival curves under non-isothermal conditions, respectively. The results indicated that microbial survival parameters could be accurately estimated from multiple whole survival curves under changing lethal agent intensity. In principle, the developed algorithm can also be  applied to studies on nutrients preservation during food processing where degradation of food nutritional compounds induced by the processing is governed by certain chemical reaction kinetics.