Intelligent and Probabilistic Models for Evaluating the Release of Food Bioactive Ingredients from Carriers/Nanocarriers

Release modeling is an important issue in designing controlled release systems for food bioactive compounds and nutraceuticals in order to elucidate the underlying mechanisms. Adequate knowledge about such mechanisms leads to designing optimal controlled delivery vehicles. The conventional modeling approaches such as mathematical and mechanistic modeling represent considerable information about the release phenomena but the limiting factor in their applicability is the poor possibility of generalization of the curve fitting results. In this study, intelligent and probabilistic models for evaluating the release of food bioactive ingredients from carriers/nanocarriers are reviewed. Stochastic methods such as Monte Carlo (MC) and cellular automata (CA) are developed based on the idea of existing random fluctuations in the release process that can be an alternative to overcome the limitations of the conventional methods. Artificial neural networks (ANNs) are also very efficient when a nonlinear relationship exists between release profiles and formulation and process factors. The release process can be optimized using adaptive neuro-fuzzy inference systems (ANFIS) and genetic algorithms (GAs). In this study, theoretical background and fundamental operating principles of probabilistic methods (MC and CA) and artificial intelligence-based methods (ANNs, ANFIS, GAs) in the field of food bioactive release from carriers/nanocarriers are introduced. Also, some application examples for utilization of such methods are briefly discussed to preset a state of the art for practical application of these models in future food-related applications.