Towards intelligent food safety: Machine learning approaches for aflatoxin detection and risk prediction

Aflatoxins pose a grave threat with potentially devastating health effects that go unnoticed in our everyday food supply, especially foods such as peanuts, maize, and spices, particularly in the tropics and sub-tropics, where climatic conditions are conducive to their production. Existing methodologies for aflatoxin determination remain costly and time-consuming, preventing their implementation on a practical level for fast and widespread deployment through real-time monitoring. This review provides a landmark, integrative perspective of how artificial intelligence (AI) in its numerous forms has advanced aflatoxin detection and quantification across agricultural systems. In particular, this review considers the integration of food science and AI by understanding the application of supervised, unsupervised, and reinforcement learning to spectral, image, and behavioral data analysis used to inform aflatoxin detection and quantification. Combined with state-of-the-art applications such as AI smartphone-based diagnostics, clever storage systems, and deep learning models for image analysis, this review examines various cases and evaluations of developed models, addressing critical real-world challenges such as sparse data, generalization across food matrices, and regulatory transparency. Ultimately, the review addresses the willingness to adopt evolving AI strategies and looks to the future for faster, wiser, and more accessible aflatoxin detection methods for more significant public health protection and sustainable food systems. Finally, regardless of whether you are a uniquely positioned researcher investigating the development of new models, a policymaker developing food safety regulations, an academic designing curriculum, or a scientist inquisitively exploring the next generation of food technologies, this article is a timely and convenient place to access knowledge leading toward safer, AI-powered food systems.