Machine Learning Assisted Simultaneous Estimation of Drugs in Multicomponent Formulations by Spectrophotometry

Abstract

Analysis of multi-drug formulations by spectrophotometric method is a challenging task due to mutual interference. Mathematical models reported for such analysis are applicable only where there are well separated absorption maxima of component drugs. In the present work Machine learning model has been developed for the simultaneous estimation of drugs in multi-drug pharmaceutical formulations using a custom made, comprehensive, interactive and user friendly software package, by spectrophotometry. Performance of the model was assessed by estimating the drugs in tablet formulation containing amlodipine besylate and losartan potassium. The accuracy, by recovery, was more than 98%. Intraday precision studies exhibited relative standard deviation of 1.67% for amlodipine besylate and 0.93% for losartan potassium where as in inter day precision studies the method exhibited relative standard deviation of 0.86% for amlodipine besylate and 0.77% for losartan potassium. The results indicated that machine learning model could be a promising tool for simultaneous estimation of drugs in multicomponent formulations by simple spectrophotometry.