Use of ComBase data to develop an artificial neural network model for nonthermal inactivation of Campylobacter jejuni in milk and beef and evaluation of model performance and data completeness using the acceptable prediction zones method

Abstract

ComBase is a widely used microbial modeling database. ComBase data can be used to develop and validate models and to test novel modeling methods like artificial neural networks (ANN) and acceptable prediction zones (APZ), which have been shown to outperform traditional methods. Here, ComBase data were used to evaluate the ANN and APZ methods for modeling nonthermal inactivation of Campylobacter jejuni in milk and beef as a function of time, temperature (À20, 1, 10, 20, 30, and 40 C), and strain (18177, ATCC 29428). Four ANN were developed using Excel and Neu-ralTools, and the best-performing was a general regression neural network (GRNN) whose performance and data completeness were evaluated using the APZ method.Relative variable impacts in the GRNN model were 42.5%, 31.5%, 20.1%, and 5.9% for time, temperature, food, and strain, respectively. Nonthermal inactivation of C. jejuni was faster and greater at ambient than at cold temperatures and in milk than in beef except at 1 C where it was similar. The proportion of residuals in the APZ (pAPZ) ranged from 0.77 to 1 for individual nonthermal inactivation curves. Although the model had acceptable performance (pAPZ ≥0.7), it failed validation because of data gaps like one instead of four replicates per combination of independent variables and no data at À10 C. Thus, these and other data gaps identified need to be filled before the model can be used with confidence to predict behavior of C. jejuni in milk and beef. Nonetheless, results indicated that ANN and APZ methods can be used to model data for nonthermal inactivation of C. jejuni in food.

Longitudinal Study for the Detection and Quantification of Campylobacter spp. in Dairy Cows during Milking and in the Dairy Farm Environment

“…The minimum temperature was used for the analysis since the teat swab samples were taken early in the morning. Previous in vitro studies demonstrated a slower inactivation of C. jejuni by oxygen at cooler temperatures [ 37 , 38 , 39 ]. This was not observed in our study, as all classes of results occurred in the same temperature range.…”

Modeling the survival of Campylobacter jejuni in raw milk considering the viable but non‐culturable cells (VBNC)

“…C. jejuni only grows under special conditions, that is, at high temperatures and low oxygen levels, and in general, its non‐thermal inactivation by oxygen is slower at cold temperatures (Boleratz & Oscar, 2022; Olofsson et al, 2015; Yoon et al, 2004). Additionally, Campylobacter spp.…”

“…Only a few studies investigated the survival of different strains of Campylobacter spp. in raw, skimmed or unpasteurized milk at distinct temperatures (Boleratz & Oscar, 2022; Christopher et al, 1982; Doyle & Roman, 1982; Jaakkonen et al, 2020; Wulsten et al, 2020), and only one investigated the possibility of survival of Campylobacter spp. in VBNC state in raw milk (Wulsten et al, 2020).…”

“…The high sensitivity of C. jejuni to oxygen and other environmental stressors (e.g., pH, temperature) (Al‐Qadiri et al, 2015; Christopher et al, 1982; Doyle & Roman, 1982; Kim et al, 2017) together with the limitation of available quantitative detection methods not only makes C. jejuni difficult to be cultured and enumerated, but also leads to highly variable results. Therefore, data for model development and validation are still limited in this area and to our knowledge, just one predictive mathematical model describing the survival of C. jejun i in milk have been developed so far, but the authors did not take into account the IPIU data, but just the CFU data (Boleratz & Oscar, 2022).…”

Artificial Intelligence Models for Zoonotic Pathogens: A Survey

“…In particular, Logistic Regression (Cox 1958 [ 2 ]) and Random Forest (Ho 1995 [ 3 ], Breiman 2001 [ 4 ]) are widely used for modeling and drawing useful inferences about zoonotic diseases and their transmission (Ntampaka et al, 2021 [ 5 ], Kiambi et al, 2020 [ 6 ], Acharya et al, 2019 [ 7 ]). More recently, the effectiveness of artificial neural networks in modeling zoonotic diseases and their causes have also been demonstrated in a number of studies (Boleratz and Oscar 2022 [ 8 ], ZareBidaki et al, 2022 [ 9 ], Denholm et al, 2020 [ 10 ]).…”