A Mathematical Model Based on an Adaptive Neuro-Fuzzy Inference System for Matrixes Including Indomethacin

Abstract

This study is concerned about prediction of dissolution rate of Insoluble drugs from solid dispersion (SD) polymer matrixes by an Adaptive Neuro-Fuzzy Inference System (ANFIS). Polyethylene Glycols (PEGs) as the SD with different molecular weights were provided and dissolution rate of indomethacin (IND) was obtained experimentally. A USP dissolution method was used to monitor the dissolution profiles of matrixes. The numbers of rules were trained in a systematic procedure using the experimental data. Comparison of IND dissolution rate from different matrixes, Area under the Curve (AUC) of absorbance vs. time diagrams in the first 25 min for 72 different samples was determined. Results show a high correlation between observed and predicted data (r2=0.85). The calculated root mean square error for the results of the ANFIS model is equal to 1.02. The index of area AUC in the first 25 min is more repeatable. It seems that the model has practical value and different ratios of polymer for the desired dissolution rate can be predicted or having different polymer ratios in the matrix can predict the dissolution rate of IND. this method can be suggested for other pharmaceuticals formulations to save time and money to achieve the best formula.