Sustained modelling ability of artificial neural networks in the analysis of two pharmaceuticals (dextropropoxyphene and dipyrone) present in unequal concentrations

Abstract

An improvement is presented on the simultaneous determination of two active ingredients present in unequal concentrations in injections. The analysis was carried out with spectrophotometric data and non-linear multivariate calibration methods, in particular artificial neural networks (ANNs). The presence of non-linearities caused by the major analyte concentrations which deviate from Beer's law was confirmed by plotting actual vs. predicted concentrations, and observing curvatures in the residuals for the estimated concentrations with linear methods. Mixtures of dextropropoxyphene and dipyrone have been analysed by using linear and non-linear partial least-squares (PLS and NPLSs) and ANNs. Notwithstanding the high degree of spectral overlap and the occurrence of non-linearities, rapid and simultaneous analysis has been achieved, with reasonably good accuracy and precision. A commercial sample was analysed by using the present methodology, and the obtained results show reasonably good agreement with those obtained by using high-performance liquid chromatography (HPLC) and a UV-spectrophotometric comparative methods.

Enhanced application of square wave voltammetry with glassy carbon electrode coupled to multivariate calibration tools for the determination of B6 and B12 vitamins in pharmaceutical preparations

“…Intrinsically non-linear calibration techniques such as artificial neural networks (ANNs) [32,33] are applicable in the latter cases. Recently we have reported the use of these kind of chemometrics tools in solving complex pharmaceutical systems with a high improvement of the results previously obtained by applying linear methods as PLS-1 [34,35].…”

Artificial Neural Networks: Theoretical Background and Pharmaceutical Applications: A Review

“…Besides the application of ANNs in assessment of purity, stability, and quantification of different crystalline forms of a drug, new easy and fast ways of acquiring data obtained with reliable instrumental techniques have allowed neural networks to be widely used for the quantification of medicinal products (50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61).…”

“…Based on the information compiled in Table 2, it can be concluded that compared to conventionally analytical methods, spectroscopic techniques are the most frequently applied in simultaneous determination of two active constituents in multicomponent, commercial medicinal products (54)(55)(56)(57)(58)(59). The great advantages of these methods, especially those using the IR range, over other analytical techniques are the simplicity of measurement, short time of analysis, and capacity for automation of the analytical procedure.…”

Determination of Paracetamol and Orphenadrine Citrate in Pharmaceutical Tablets by Modeling of Spectrophotometric Data Using Partial Least-Squares and Artificial Neural Networks

“…9) It has also proved useful for resolving mixtures of analytes giving deviate signal form Beer and Lambert's Law in spectrophotometric techniques. 10) Paracetamol (4-acetamidophenol) is an eŠective 30 2 12 20 2 2 10 2 13 10 0 3 20 1 14 15 0 4 20 3 15 20 0 5 25 1.5 16 25 0 6 25 2.5 17 30 0 7 1 5 1 . 5 1 8 0 1 8 1 5 2 .…”