Raman imaging and statistical methods for analysis various type of human brain tumors and breast cancers
Monika Kopeć, Maciej Błaszczyk, Maciej Radek, Halina Abramczyk
Abstract
Spectroscopic methods provide information on the spatial localization of biochemical components based on the analysis of vibrational spectra. Raman spectroscopy and Raman imaging can be used to analyze various types of human brain tumors and breast cancers. The objective of this study is to evaluate the Raman biomarkers to distinguish tumor types by Raman spectroscopy and Raman imaging. We have demonstrated that bands characteristic for carotenoids (1156 cm −1 , 1520 cm −1 ), proteins (1004 cm −1 ), fatty acids (1444 cm −1 , 1655 cm −1 ) and cytochrome (1585 cm −1 ) can be used as universal biomarkers to assess aggressiveness of human brain tumors. The sensitivity and specificity obtained from PLS-DA have been over 73%. Only for gliosarcoma WHO IV the specificity is lower and takes equal 50%. The presented results confirm clinical potential of Raman spectroscopy in oncological diagnostics.