K K Peh1, C P Lim, S S Quek, K H Khoh. 1. School of Pharmaceutical Sciences, University of Science Malaysia, Penang. kkpeh@usm.my
Abstract
PURPOSE: To use artificial neural networks for predicting dissolution profiles of matrix-controlled release theophylline pellet preparation, and to evaluate the network performance by comparing the predicted dissolution profiles with those obtained from physical experiments using similarity factor. METHODS: The Multi-Layered Perceptron (MLP) neural network was used to predict the dissolution profiles of theophylline pellets containing different ratios of microcrystalline cellulose (MCC) and glyceryl monostearate (GMS). The concepts of leave-one-out as well as a time-point by time-point estimation basis were used to predict the rate of drug release for each matrix ratio. All the data were used for training, except for one set which was selected to compare with the predicted output. The closeness between the predicted and the reference dissolution profiles was investigated using similarity factor (f2). RESULTS: The f2 values were all above 60, indicating that the predicted dissolution profiles were closely similar to the dissolution profiles obtained from physical experiments. CONCLUSION: The MLP network could be used as a model for predicting the dissolution profiles of matrix-controlled release theophylline pellet preparation in product development.
PURPOSE: To use artificial neural networks for predicting dissolution profiles of matrix-controlled release theophylline pellet preparation, and to evaluate the network performance by comparing the predicted dissolution profiles with those obtained from physical experiments using similarity factor. METHODS: The Multi-Layered Perceptron (MLP) neural network was used to predict the dissolution profiles of theophylline pellets containing different ratios of microcrystalline cellulose (MCC) and glyceryl monostearate (GMS). The concepts of leave-one-out as well as a time-point by time-point estimation basis were used to predict the rate of drug release for each matrix ratio. All the data were used for training, except for one set which was selected to compare with the predicted output. The closeness between the predicted and the reference dissolution profiles was investigated using similarity factor (f2). RESULTS: The f2 values were all above 60, indicating that the predicted dissolution profiles were closely similar to the dissolution profiles obtained from physical experiments. CONCLUSION: The MLP network could be used as a model for predicting the dissolution profiles of matrix-controlled release theophylline pellet preparation in product development.
Authors: H Malinowski; P Marroum; V R Uppoor; W Gillespie; H Y Ahn; P Lockwood; J Henderson; R Baweja; M Hossain; N Fleischer; L Tillman; A Hussain; V Shah; A Dorantes; R Zhu; H Sun; K Kumi; S Machado; V Tammara; T E Ong-Chen; H Mahayni; L Lesko; R Williams Journal: Adv Exp Med Biol Date: 1997 Impact factor: 2.622
Authors: Dorián László Galata; Attila Farkas; Zsófia Könyves; Lilla Alexandra Mészáros; Edina Szabó; István Csontos; Andrea Pálos; György Marosi; Zsombor Kristóf Nagy; Brigitta Nagy Journal: Pharmaceutics Date: 2019-08-09 Impact factor: 6.321