Modeling in Food Microbiology

From Predictive Microbiology to Exposure Assessment

Jeanne-Marie Membré, Vasilis Valdramidis

• Publisher: ISTE Press - Elsevier
• Year: 2016
• Print ISBN: 9781785481550
• eBook ISBN: 9780081009819
• Topic: Technology & Engineering
• Subtopic: Medical Microbiology & Parasitology
• Index: Technology & Engineering

Predictive microbiology primarily deals with the quantitative assessment of microbial responses at a macroscopic or microscopic level, but also involves the estimation of how likely an individual or population is to be exposed to a microbial hazard.This book provides an overview of the major literature in the area of predictive microbiology, with a special focus on food. The authors tackle issues related to modeling approaches and their applications in both microbial spoilage and safety.Food spoilage is presented through applications of best-before-date determination and commercial sterility. Food safety is presented through applications of risk-based safety management. The different modeling aspects are introduced through probabilistic and stochastic approaches, including model and data uncertainty, but also biological variability. - Features an extensive review of modelling terminology - Presents examples of all available microbial models (i.e., growth, inactivation, growth/no growth) and applicable software - Revisits all statistical aspects related to exposure assessment - Describes realistic examples of implementing microbial spoilage and safety modeling approaches

1

Predictive Microbiology

Vasilis Valdramidis

Abstract:

Overall, the models developed in predictive microbiology aim at the quantification of the effects of intrinsic, extrinsic and/or processing factors on the resulting microbial proliferation in food products or food model systems, for example, a buffer system. These models rely on the possibility to interpolate the resulting microbial proliferation for combinations that are not only originally examined, but also included in the range of the experiment design. As such, predictive microbiology can be considered as a powerful tool to investigate and summarize succinctly the effect of varying conditions (within food formulation and processing) on the microbial ecology. A historical perspective of modeling developments in the area of predictive microbiology was presented by Mafart in 2005. According to that, the first developments date back to 1920s when heat resistance of microorganisms was described by either the Arrhenius equation or the Bigelow model. Nevertheless, the principles and goals of the discipline appeared much later, at the beginning of the 1990s, followed by the development and description of microbial models and the generation of relevant databases and other software tools.

Keywords

Categorical modeling

Kinetic responses

Modeling cycle

Model validation

Predictive microbiology

A priori microbiological knowledge

Regression analysis

Structural characteristics

1.1 Introduction

Overall, the models developed in predictive microbiology aim at the quantification of the effects of intrinsic, extrinsic and/or processing factors on the resulting microbial proliferation in food products or food model systems, for example, a buffer system (e.g. [WHI 95]). These models rely on the possibility to interpolate the resulting microbial proliferation for combinations that are not only originally examined, but also included in the range of the experiment design. As such, predictive microbiology can be considered as a powerful tool to investigate and summarize succinctly the effect of varying conditions (within food formulation and processing) on the microbial ecology. A historical perspective of modeling developments in the area of predictive microbiology was presented by Mafart in 2005 [MAF 05]. According to that, the first developments date back to 1920s when heat resistance of microorganisms was described by either the Arrhenius equation [ARR 89] or the Bigelow model [BIG 21]. Nevertheless, the principles and goals of the discipline appeared much later, at the beginning of the 1990s, followed by the development and description of microbial models and the generation of relevant databases and other software tools.

An outline of the most recent developments and applications of predictive microbiology as well as some representative published examples is given as follows:

– studies on the assessment of shelf-life (e.g. [XIA 14, HUC 13]);

– applications for process design and optimization (e.g. [VAL 07, KAT 10]);

– development of exposure assessment applications (e.g. [TEN 15, MEM 09, PUJ 15]);

– integration of modeling approaches in systems biology (e.g. [BRU 08, VAN 13]).

In the following sections of the chapter, the main principles of predictive microbiology are discussed by providing an overview and interpretation of terms used in the area as well as a classification and a description of the procedure to develop models. Finally, examples of software developments and further literature studies are provided.

1.2 Terminology

A selection of the main terminology in the discipline of predictive microbiology is provided:

– variables: divided in dependent and independent variables. Dependent variables describe a certain response (e.g. microbial population) in relation to some independent variables (e.g. temperature and pH). The independent variables could be also named factors;

– regression analysis: statistical process for estimating the relationship between independent variables and dependent variable, and estimate the model parameters. In predictive microbiology this could involve food extrinsic (e.g. temperature) and intrinsic (e.g. pH and a_w) variables and their relationship;

– prediction: the use of models to forecast responses;

– parameters: set of measurable values which are estimated when the model is solved; they provide the final modeling structure;

– simulation: the use of models to describe responses for a set of predefined variables/factors over time by the use of nominal values. Simulations can be used to perform predictions;

– validation: the process of deciding whether the numerical results quantifying hypothesized relationships between variables, obtained from regression analysis, are acceptable as descriptions of the data. It is an imperative step before using a new modeling structure for other microbiological applications (e.g. performing microbial predictions in food).

The current terminology will help the reader to understand the model development steps and also topics related to the type of models and applications that are available in the literature.

1.3 Classification

Numerous modeling approaches have been published the past years in the literature. These models aim at addressing different issues or interpreting different microbiological phenomena. It is therefore essential for the reader that he/she is first introduced in the area by reviewing how models can be classified based on a number of criteria. These read as follows:

– Structural characteristics: based on these characteristics, the models are described as (1) white box or mechanistic (physical) models which are constructed...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Introduction
• 1: Predictive Microbiology
• 2: Quantifying Microbial Propagation
• 3: Modeling Microbial Responses: Application to Food Spoilage
• 4: Modeling Microbial Responses: Application to Food Safety
• Conclusion
• List of Authors
• Index