Paralelna faktorska analiza fluorescentnih svojstava višekomponentnih sistema

Abstract

Kombinacija optičkih spektroskopskih tehnika i odgovarajućih metoda za analizu imodeliranje spektralnih podataka se pokazala kao veoma obećavajuća tehnika zakarakterizaciju i analizu organskih i neorganskih kompleksnih sistema. Razumevanjefizičkih fenomena koji se javljaju u kompleksnim sistemima predstavlja bitan preduslovza primenu spektroskopskih tehnika u različitim oblastima nauke. U ovoj tezi su izmerenefluorescentne karakteristike četiri kompleksna sistema: tri organska (rastvor vode iaminokiselina, med i tkivo dojke) i jedan neorganski (kompleksni sistem fosfora na baziretkih zemalja). Dobijeni fluorescentni podaci su modelirani primenom paralelnefaktorske analize. Izgrađeni modeli su omogućili definisanje fluorescirajućih komponenti(fluorofora) prisutnih u sistemima, pri čemu je model kao rezultat dao čiste spektre svakepojedinačne fluorofore, kao i njihove relativne koncentracije u svakom od uzorakasistema. U neorganskom sistemu i rastvoru aminokiselina broj fluorofora i njihovekoncentracije su unapred bile poznate i na taj način je testirana uspešnost modela kao injegova sposobnost predikcije prisustva i koncentracija fluorofora u nepoznatimuzorcima. Pokazano je da je dobijeni model veoma uspešan, sa odstupanjem od samo0.036% od stvarne koncentracije kod neorganskog sistema. Za sistem meda, model jepokazao da je u sistemu prisutno 6 fluorofora i dobijeni su čisti spektri svake od njih kaoi njihove relativne koncentracije u svakom pojedinačnom uzorku. Dalje analize spektarameda su ukazale da je na osnovu koncentracija fluorofora moguće odrediti botaničkoporeklo meda. Spektri meda su zatim analizirani metodom glavnih komponenata kao idiskriminantnom analizom parcijalno najmanjih kvadrata. Rezultati su pokazali velikuuspešnost primene metode fluorescentne spektroskopije za klasifikaciju i autentifikacijuuzoraka meda. Proračunom modela tkiva dojke je ustanovljeno da postoje 4 dominantnefluorofore. Utvrđeno je da je na osnovu njihovih koncentracija moguće sa tačnošću od100% odrediti vrstu promena na tkivu (benigno ili maligno). Rezultati su potom poređenisa rezultatima analize potpornih vektora primenjenih na neobrađene spektre, na osnovukojih je pokazano da se sa istim podacima ne može dobiti podjednako uspešan model...

The combination of optical spectroscopic techniques and appropriate methods foranalysis and modeling of spectral data proved to be a very promising technique for thecharacterization and analysis of organic and inorganic complex systems. Understandingof physical phenomena that occur in complex systems is an essential prerequisite for theapplication of spectroscopic techniques in various fields of science. Fluorescencecharacteristics of four complex systems: three organic (solution of water and amino acids,honey and breast tissue) and one inorganic (a complex system of phosphorus-based rareearth) were measured in this thesis. The obtained fluorescence data were modeled usingparallel factor analysis. The constructed models have enabled definition of the fluorescingcomponents (fluorophores) present in the system, wherein the model as a result gave purespectra of the individual fluorophore as well as their relative concentrations in eachsample of the system. In the inorganic system and solution of amino acids, the number offluorophores and their concentrations were known in advance, and thus the performanceof the model was tested as well as its ability to predict the presence and concentrations offluorophores in the unknown samples. It was shown that the resulting model is verysuccessful, with a deviation of only 0.036% of the actual concentration in inorganicsystems. For the system of honey, the model showed that 6 fluorophores are present inthe system and their pure spectra were obtained, as well as their relative concentrationsin each sample. Further analysis of the spectra of honey have demonstrated that based onthe concentrations of the fluorophores the botanical origin of honey can be determined.Spectra of honey were then analyzed using principal component analysis and discriminantpartial least squares. Results have shown great success in application of fluorescencespectroscopy methods for classification and authentication of honey samples. Based oncalculated model of the breast tissue it was determined that there are four dominantfluorophores. It was estabilshed that based on their concentrations it possible to determinethe type of change in the tissue (benign or malignant) with an accuracy of 100%. Theresults were then compared with the results of support vector machine analysis applied tothe raw spectra, from which it was shown that using the same data successful model cannot be obtained (classification error 35.71%). However, it has been shown that with...