Marek Wesołowski

Statystyczne techniki wielowymiarowe w analizie farmaceutycznej
2006-05-16

Statistical multivariate techniques in pharmaceutical analysis. In the work, selected examples of application of multivariate methods in the analysis of huge measurement databases, obtained by means of modern and fully automated instrumental techniques of analysis, are described. After short introduction, in which the most frequently used chemometric techniques in the pharmaceutical analysis - principal component analysis (PCA), cluster analysis (CA) and artificial neural networks (ANN), were presented, the usefulness of these techniques in the simultaneous determination of whewellite, weddellite and uric acid in human urinary calculi; classification of wines by their geographic origin, taking into consideration the interpretation of concentration of selected elements in wine samples; and in the optimization of the chemical composition of theophylline tablets in order to obtain desirable release profiles of the drug substance from drug formulation, is also given. Presented examples indicate, that by gathering together the elements of informatics and the chemical analysis, mentioned above multivariate techniques appeared to be extremely useful in the professional work of pharmacists, connected with the interpretation of the measurement data and the extraction from them the maximum of useful information.