Application of data-intelligence algorithms for modeling the compaction performance of new pharmaceutical excipients

Abstract

Pharmaceutical excipient development is an extensive process requiring a series of pre-formulation studies to evaluate their performance. The present study compares the conventional compaction and compression pre-formulation studies with artificial intelligence (AI) modeling to predict the performances of thermally and chemically modified starches obtained from Livingstone potato. The native starch was modified by three methods, and we obtained the following starches; pregelatinized starch (PS), ethanol dehydrated pregelatinized starch (ES), and acid hydrolyzed starch (AS). Microcrystalline cellulose (Avicel® PH101) was employed as a reference since its use in tablet direct compression has been established. The role of compaction pressure on the degree of volume reduction of the tablets was studied using Kawakita and Heckel models which highlighted that when the starch is modified by pregelatinization followed by ethanol dehydration, and/and or acid hydrolysis modification, a directly compressible starch can be obtained that can plastically deform. The data-intelligence results indicated the reliability of the AI-based models over the linear models. Hence, the comparative results demonstrated that the Adaptive neuro-fuzzy inference system (ANFIS) outperformed the other two models in modeling the performance of all of the four excipients with considerable performance accuracy. The compressibility indices have shown matching characteristics of AS and ES to Avicel® PH101 in terms of direct compressibility potential than PS. Moreover, the data intelligence modeling demonstrates the reliability and satisfactory of ANFIS as a hybrid model over the other two models with increased performance skills in modeling the compaction properties of these novel pharmaceutical excipients.