From Microbial Transformation to Machine Learning Optimization: Health Benefits and Technological Advances in Fermented Fruits and Vegetables

Fermented fruits and vegetables (FFVs) have attracted attention due to their remarkable ability for preventing chronic diseases. This review focuses on the mechanisms of microbial transformation in enhancing the health benefits of FFVs, and introduces the potential of machine learning (ML) in improving the production efficiency and product quality of FFVs. Fermentation strains represented by lactic acid bacteria, yeasts, and acetic acid bacteria play a protective role not only via direct colonization but also through inhibiting pathogens, biotransforming active substances, and promoting the degradation of anti-nutritional factors. Compared to unfermented counterparts, FFVs exhibit stronger antioxidant activity, angiotensin-converting enzyme inhibition, higher bioavailability, and intestinal microbiota modulation, aiding in managing obesity, diabetes, cardiovascular diseases, and intestinal disorders. However, safety concerns like fermentation strain risks, biogenic amines, nitrites, and high salt content must be addressed. The integration of ML with multi-omics enables predictive modeling of fermentation results, precise strain selection, and quality control, thereby achieving accurate monitoring of dynamic fermentation process. The focus of future work should be on establishing large-scale databases that integrate multi-omics and synthetic biology to further improving fermentation efficiency and safety while developing advanced FFVs.