Application of an Artificial Neural Network for Design of Sustained-Release Matrix Tablets Containing Vaccinium Myrtillus Leaf Powder Extract

Abstract

Context: Vaccinium myrtillus leaf extracts are promising source of natural remedies for diabetes mellitus Type 2 management and prevention. Aim: The aim of this study was to design the sustained-release matrix tablets containing V. myrtillus leaf powder extract with the application of an artificial neural network (ANN). Methods and Materials: The amounts of Methocel K4M, Methocel K100LV, and Eudragit L100 were used as input factors affecting the release from matrix tablets. Each input factor was varied on three levels according to Box–Behnken design. The in vitro percent release at time points of 2, 8, and 16 h were used as output data in training, testing, and validating the neural network. The software Matlab was used to create an ANN and the number of nodes in the hidden layer was selected based on trial and error approach to develop a model with the best predictive ability. Results: The multilayer perceptron with one hidden layer was constructed. The network with nine nodes in the hidden layer was used to simulate in vitro release from hypothetical formulations and the matrix forming agent ratio was selected by the brute-force method. The dissolution profile for the selected formulation of matrix tablets was studied. The evaluation of the release kinetics and mechanism indicated a coupling of the diffusion and erosion as release mechanisms. Conclusions: ANNs can be successfully applied to develop herbal sustained release matrix tablets.