A Raman-based serum constituents' analysis for gastric cancer diagnosis: In vitro study

The importance of blood serum testing in most medical diagnosis has led researchers to seek for a reliable analysis method. Raman spectroscopy is a potential tool for probing serum components due to its real-time and non-destructive measurements without need to any additional reagents. So, Raman spectroscopy for the analysis and discrimination of human serum of healthy and gastric cancer subjects is investigated in this work. In order to find the correlation between Raman spectra and enzymatic test results of glucose, cholesterol, HDL, LDL and triglycerides of serum samples, the partial least squares regression (PLSR) method is utilized. Moreover, the Raman spectroscopy is used to distinguish between serum samples of healthy people and gastric cancer patients using partial least square discriminant analysis (PLS-DA) method. Correlation results reveal that for all serum components, the correlation coefficients between Raman spectra and all enzymatic test results are above 94% significantly. Discrimination results of healthy subjects and gastric cancer patients show that 87.5 +/- 2.5% of healthy and gastric cancer subjects are diagnosed properly. Our preliminary results can confirm the Raman method in analysis of serum and also as a diagnostic tool in screening of gastric cancer.