Optimization of Salbutamol Sulfate Dissolution from Sustained Release Matrix Formulations Using an Artificial Neural...

Area of Science:

• Pharmaceutical Science
• Computational Chemistry
• Drug Delivery Systems

Background:

• Hydrophilic matrix formulations are crucial for controlled drug release.
• Optimizing drug release profiles requires careful selection of excipients and their concentrations.
• Artificial neural networks (ANNs) offer a powerful tool for complex system modeling.

Purpose of the Study:

• To utilize artificial neural networks (ANNs) for optimizing the in vitro release of salbutamol sulfate from hydrophilic matrix formulations.
• To develop a predictive model for formulation development and optimization.
• To assess the utility of ANNs in achieving desired drug release characteristics.

Main Methods:

• Formulations were designed using a central composite design, varying Methocel K100M, xanthan gum, Carbopol 974P, and Surelease.
• In vitro dissolution data at six time points served as target data for training the ANN.
• A multi-layer perceptron model was developed and optimized using Matlab, determining the optimal number of hidden layer nodes.

Main Results:

• A neural network with nine hidden layer nodes demonstrated optimal predictive ability for formulation optimization.
• Simulations confirmed the usability and predictive accuracy of the developed ANN model.
• The optimized ANN successfully guided the development of a formulation with predicted desirable release characteristics, which was validated experimentally.

Conclusions:

• Artificial neural networks provide a viable and effective tool for optimizing pharmaceutical formulations.
• ANNs can accurately predict and guide the development of drug delivery systems with specific performance characteristics.
• This study highlights the potential of ANNs in accelerating and improving the formulation development process.