Formulation Optimization of Low Bioavailable Drug Loaded Alginate Microparticles Using Artificial Neural Networks

In the present study, sodium alginate microparticle for oral delivery of furosemide was designed whether the encapsulation into microparticles might improve the oral absorp-tion of this potent loop diuretic. We described preparation of microspheres based on ionotropic gelation method and characterized its physicochemical properties.To acquire an optimum formulation, a Generalized Regression Neural Networks (GRNN) and a Multi-Layer Perceptron (MLP) were employed. The drug loaded formulation parame-ters were the input vectors of the GRNN and included the amount of polymer, cross linked agent, volume of external and internal phases. The microparticles drug loading, size and in vitro drug release constitute the output vector of each network. In this way, GRNN and MLP were trained to investigate the functional influence of input variables on the output responses. The results demonstrated that GRNN is promising in providing better solutions for optimization of drug delivery system formulation.The obtained op-timum formulation showed a narrow size distribution on an average diameter of 700 ± 50 m and drug loading of more than 75%. The drug release profile illustrates a sus-tained released pattern and released percent of about 36 % in 2 hour. In vitro drug re