The management of mycotoxin contamination in the supply chain is continuously evolving in response to growing knowledge about mycotoxins, shifting factors that influence mycotoxin occurrence, and ongoing technological developments. One of the technological developments is the potential for using artificial intelligence (AI) in mycotoxin management. AI can be used in various fields of mycotoxin management, including for predictive modelling of mycotoxins and for analytical detection and analyses. This review aimed to investigate the state-of-the-art of the use of AI for mycotoxin management. This review focuses on (1) predictive models for the presence of mycotoxins in commodities at both pre-harvest and post-harvest levels and (2) the detection of mycotoxins in samples by processing large datasets resulting from imaging data or chemical analyses of the sample. A systematic review was conducted, resulting in a total of 70 relevant references, including 15 references focusing on mycotoxin prediction models and 54 references focusing on mycotoxin detection, ranging from imaging to chemical analysis, and including relevant reviews. The AI applications and the most popular AI algorithms are presented. As shown by this review, AI is able to improve mycotoxin prediction models both at pre- and post-harvest levels and makes the emergence of non-invasive and fast detection methods such as imaging detection or electronic noses possible. A major challenge remains in the applicability and scalability of AI models to practical settings.

Agricultural commodities used for feed and food production are frequently contaminated with mycotoxins, such as Aflatoxin B1 (AFB1). In Europe, both the government and companies have monitoring programs in place for the presence of AFB1. With limited resources and following risk-based monitoring as prescribed in EU Regulation 2017/625, these monitoring programs focus on batches with the highest probability of being contaminated. This study explored the use of machine learning algorithms (ML) to design risk-based monitoring programs for AFB1 in feed products, considering both monitoring cost and model performance. Historical monitoring data for the presence of AFB1 in feed products (2005–2018; 5605 records in total) were used. Four different ML algorithms, including Decision tree, Logistic regression, Support vector machine and Extreme gradient boosting (XGB), were applied and compared to predict the high-risk feed batches to be considered for further AFB1 sampling and analysis. The monitoring cost included the cost of: sampling and analysis, disease burden, storage, and of recalling and destroying contaminated feed batches. The ML algorithms were able to predict the high-risk batches, with an AUC, recall, and accuracy higher than 0.8, 0.6, and 0.9, respectively. The XGB algorithm outperformed the other three investigated ML. Its incorporation would result into up to 96% reduction in monitoring cost in 2016–2018, as compared to the official monitoring program. The proposed approach for designing risk based monitoring programs can support authorities and industries to reduce the monitoring cost for other food safety hazards as well.

Historical data on food safety monitoring often serve as an information source in designing monitoring plans. However, such data are often unbalanced: a small fraction of the dataset refers to food safety hazards that are present in high concentrations (representing commodity batches with a high risk of being contaminated, the positives) and a high fraction of the dataset refers to food safety hazards that are present in low concentrations (representing commodity batches with a low risk of being contaminated, the negatives). Such unbalanced datasets complicate modeling to predict the probability of contamination of commodity batches. This study proposes a weighted Bayesian network (WBN) classifier to improve the model prediction accuracy for the presence of food and feed safety hazards using unbalanced monitoring data, specifically for the presence of heavy metals in feed. Applying different weight values resulted in different classification accuracies for each involved class; the optimal weight value was defined as the value that yielded the most effective monitoring plan, that is, identifying the highest percentage of contaminated feed batches. Results showed that the Bayesian network classifier resulted in a large difference between the classification accuracy of positive samples (20%) and negative samples (99%). With the WBN approach, the classification accuracy of positive samples and negative samples were both around 80%, and the monitoring effectiveness increased from 31% to 80% for pre‐set sample size of 3000. Results of this study can be used to improve the effectiveness of monitoring various food safety hazards in food and feed.

Fusarium head blight (FHB) is a plant disease caused by various species of the Fusarium fungus. One of the major concerns associated with Fusarium spp. is their ability to produce mycotoxins. Mycotoxin contamination in small grain cereals is a risk to human and animal health and leads to major economic losses. A reliable site-specific precise Fusarium spp. infection early warning model is, therefore, needed to ensure food and feed safety by the early detection of contamination hotspots, enabling effective and efficient fungicide applications, and providing FHB prevention management advice. Such precision farming techniques contribute to environmentally friendly production and sustainable agriculture. This study developed a predictive model, Sága, for on-site FHB detection in wheat using imaging spectroscopy and deep learning. Data were collected from an experimental field in 2021 including (1) an experimental field inoculated with Fusarium spp. (52.5 m × 3 m) and (2) a control field (52.5 m × 3 m) not inoculated with Fusarium spp. and sprayed with fungicides. Imaging spectroscopy data (hyperspectral images) were collected from both the experimental and control fields with the ground truth of Fusarium-infected ear and healthy ear, respectively. Deep learning approaches (pretrained YOLOv5 and DeepMAC on Global Wheat Head Detection (GWHD) dataset) were used to segment wheat ears and XGBoost was used to analyze the hyperspectral information related to the wheat ears and make predictions of Fusarium-infected wheat ear and healthy wheat ear. The results showed that deep learning methods can automatically detect and segment the ears of wheat by applying pretrained models. The predictive model can accurately detect infected areas in a wheat field, achieving mean accuracy and F1 scores exceeding 89%. The proposed model, Sága, could facilitate the early detection of Fusarium spp. to increase the fungicide use efficiency and limit mycotoxin contamination.

Weather conditions and agronomical factors are known to affect Fusarium spp. growth and ultimately deoxynivalenol (DON) contamination in oat. This study aimed to develop predictive models for the contamination of spring oat at harvest with DON on a regional basis in Sweden using machine-learning algorithms. Three models were developed as regional risk-assessment tools for farmers, crop collectors, and food safety inspectors, respectively. Data included: weather data from different oat growing periods, agronomical data, site-specific data, and DON contamination data from the previous year. Results showed that: (1) RF models were able to predict DON contamination at harvest with a total classification accuracy of minimal 0.72; (2) good predictions could already be made in June; (3) rainfall, relative humidity, and wind speed in different oat growing stages, followed by crop variety and elevation were the most important features for predicting DON contamination in spring oats at harvest.