Development and optimization of antioxidant polyherbal cream using artificial neural network aided response surface methodology

Abstract

Artificial Neural Networks, has become phenomenal in drug modeling, in predicting experimental outcomes and in optimization studies during product development. This research seeks to use artificial neural networks (ANN) to optimize a novel formulation of Polyherbal face cream composed of Cymbopogon citratus, Hibiscus sabdariffa, and Ocimum gratissimum extracts. Central composite design was used to develop a framework for the products to be studied delineating the independent variables to be concentration of oil phase, and that of the emulsifying agents used in formulation. The dependent variables/responses are viscosity, spreadability index, and particle size of the cream. The responses gotten were then optimized using ANN models. The validity of the statistical models used for predicting the observed responses were confirmed by carrying out three confirmation experimental runs at the identified optimum conditions. The viscosity of the formulated cream decreased with increase in the amount of oil phase. However, there was a stronger correlation with amount of emulsifying phase and viscosity of the cream as increased in emulsifier product. The particle sizes did not vary greatly among the various formulations irrespective of the concentration of plant extract. Formulation of Polyherbal face cream composed of Cymbopogon citratus, Hibiscus sabdariffa, and Ocimum gratissimum extracts were shown to have significant antioxidant activity. Using ANN modeling, a prediction of an optimized formulation was made and was shown to have comparable results with the predicted values.