Alginate gel scaffolds are biocompatible and biodegradable materials that have been used in a variety of tissue engineering applications. The porosity of alginate gel scaffolds is an important property that affects their performance. However, it is difficult to predict the porosity of alginate gel scaffolds accurately. In this study, a GA-coupled ANN model was developed to predict the porosity of alginate gel scaffolds. The model was trained on a dataset of 107 scaffolds with known porosities. The model was able to achieve a mean absolute error of 0.13, which suggests that it is able to accurately predict the porosity of alginate gel scaffolds. The alginate scaffold was fabricated by a microfluidic technique using a syringe pump and a flow device. The crosslinker solution was poured into the Petri dish to crosslink the polymer to the gel structure. The Archimedes method was used to determine the scaffold's apparent porosity. The artificial neural network has been used to model the porosity of the gel scaffold using the input parameters such as alginate-pluronic viscosity, surface tension, and contact angle etc. The maximum porosity was modelled to be 96.4 % using GA whereas the experimental value for the same was measured to be 92.8 ± 2 %. A 3.7% variation in the porosity was found from modelled value. To the best of our knowledge, this study is the first to develop an integrated ANN-coupled GA model to predict the maximum porosity of the gel scaffold. The result indicates that artificial intelligence has great potential for optimizing the parameters to fabricate the gel scaffold that can be used for tissue engineering applications.
Keywords: ANN; Drug release; Genetic algorithm; Porosity; Scaffold.
Copyright © 2023 Elsevier Ltd. All rights reserved.