Modeling the release behavior of limonene from the amylose nanostructures within the simulated digestive system by using adaptive-network-based fuzzy inference system (ANFIS)

The nanoencapsulation and targeted delivery of nutritional ingredients in the right location and appropriate time within the digestive tract is one of the key elements in fabricating functional food products and to deliver the bioactive components incorporated in these foodstuff. With respect to the complexities of the food network and the gastrointestinal system, it seems necessary to determine a method in order to evaluate and predict the release level of the wrapped bioactive component in the desired location and time. In this study, the release behavior of limonene from amylose nanostructures were examined considering the circumstances of the digestive system and next the adaptive-network-based fuzzy inference system was used to model the release behavior of limonene. In the design of this model, limonene concentration, amylose concentration, the time of the execution of ultrasonic stress plus the incubation time were considered as the input parameters in the circumstances of the simulated small intestine. Also, the release level was chosen as the output parameter and in order to optimize the models several membership functions together with numerous learning cycles were implemented in the form of shooting method. By the evaluation of different models, triangular membership function with 4 functions for each of the inputs was selected as the optimum model. The high value of the coefficient of determination (0.99) plus the mean square error of 0.45 in the optimized model implies the acceptable accuracy of this method to monitor the release level in the digestive circumstances. With respect to the simplicity and the acceptable accuracy of these intelligent models, they can be suggested as appropriate solutions to evaluate the controlled release behavior within the food network and the digestive system as well.