Abstract：
Recently, Chemometric calibration methods in spectrophotometric analysis are achieving significant attention in the quality control of resolving drug mixtures and pharmaceutical formulations containing two or more drugs with overlapping spectra. The simple univariate methods have been used over the last few decades and has proven to be highly efficient and easy to apply. In this study, a comparative study was performed between some univariate and multivariate methods to determine if chemometric methods can substitute univariate methods in pharmaceutical analysis. In this study, three chemometric techniques were compared to seven univariate techniques to resolve a mixture of mefenamic acid and febuxostat in their raw materials, dosage forms and spiked human plasma. Mefenamic acid and febuxostat were used together for treatment of gout. The applied chemometric methods are partial least squares (PLS), artificial neural network (ANN) and genetic algorithm partial least squares (GA-PLS), while the used univariate methods include first derivative, second derivative, ratio spectra, derivative ratio spectra, ratio subtraction, Q-Absorbance ratio and mean centering spectrophotometric methods. The ten proposed methods were found to be green, sensitive, and rapid. They are simple and did not require any pre-separation steps. The results of both univariate and multivariate approaches were statistically compared with the reported spectrophotometric methods using student\'s t test and ratio variance F-test. They were also compared with each other, using one-way analysis of variance (ANOVA). These methods were assessed and validated according to ICH guidelines. The studied drugs were analyzed in their pharmaceutical dosage forms and spiked human plasma with good recoveries using the developed methods, which qualify them for routine quality control of the studied drugs.
摘要：
最近,分光光度分析中的化学计量学校准方法在拆分包含两种或多种具有重叠光谱的药物的药物混合物和药物制剂的质量控制中获得了极大的关注。简单的单变量方法已经在过去的几十年中被使用，并且已经证明是高效且易于应用的。在这项研究中,在一些单变量和多变量方法之间进行了比较研究，以确定化学计量学方法是否可以替代药物分析中的单变量方法。在这项研究中,将三种化学计量学技术与七种单变量技术进行比较，以解决甲芬那酸和非布索坦在其原料中的混合物，剂型和加标人血浆。甲芬那酸和非布索坦一起用于治疗痛风。应用的化学计量学方法是偏最小二乘(PLS)，人工神经网络(ANN)和遗传算法偏最小二乘(GA-PLS),虽然使用的单变量方法包括一阶导数，二阶导数,比率光谱，导数比率光谱，比率减法，Q-吸光度比和平均定心分光光度法。提出的十种方法被发现是绿色的，敏感,和快速。它们是简单的并且不需要任何预分离步骤。使用studentt检验和比率方差F检验，将单变量和多变量方法的结果与报告的分光光度法进行统计学比较。他们也互相比较，使用单向方差分析(ANOVA)。这些方法根据ICH指南进行评估和验证。使用开发的方法以其药物剂型和加标的人血浆分析了研究药物，并具有良好的回收率，这使他们有资格进行所研究药物的常规质量控制。
