Back-Propagation Neural Network Model For Simultaneous Spectrophotometric Estimation Of Losartan Potassium And Hydrochlorothiazide In Tablet Dosage
The development of multivariate calibration model with back-propagation neural network using calibration sets constructed from the spectral data of pure components is proposed for the simultaneous estimation of active components, losartan potassium and hydrochlorothiazide in tablet dosage. The calibration sets were designed such that the concentrations were orthogonal and span the possible mixture space fairly evenly. The back-propagation neural network model was optimized with respect to the spectral input, training parameters and topology including transfer functions for each layer so as to yield accurate and precise estimations on model validation. The optimized model showed sufficient robustness even when the calibration sets were constructed from different set of pure spectra of components thus enabling periodical validation of model rapidly and economically. Although the components showed significant spectral overlap, the model could accurately estimate the drugs, with satisfactory precision and accuracy, in tablet dosage with no interference from excipients as indicated by the recovery study results.