Surface-Enhanced Raman Spectroscopy For Gastric Cancer And Normal Tissue

Gastric cancer is the most common malignant tumor in China of all kinds oftumors, impairing people’s health seriously. At present，the diagnosis methods ofgastric cancer mainly include X rays, gastroscope, ultrasonic etc, and the traditionalpathological examination is considered as the gold standard. While most of theexaminations are not only cause some pains and injuries to patients, but also timeconsuming. Raman spectroscopy is paid attention to its advantages of real timedetection, non-invasive and so on. But the Raman scattering section is small andRaman signals of tissue are weak. Surface-enhanced Raman spectroscopy （SERS），which can enhance about10folds than normal Raman spectra, has the very highdetection sensitivity. The purpose of this study is to detect gastric tissues and todistinguish cancerous and normal tissues with SERS.The SERS spectra of tissues, homogenate and supernatant of homogenized tissues,from26patients, are measured in this paper. The results show that the enhancementeffect, especially for supernatant of homogenized tissues, is obvious; comparing withnormal tissues, the gastric cancer tissues have higher Raman intensities at724cm^-1,1244cm^-1,1329cm^-1,1367cm^-1and1597cm^-1, but lower at828cm^-1,1002cm^-1and1128cm^-1; comparing with normal homogenate, the gastric cancer homogenate hashigher Raman intensities at623cm^-1,656cm^-1,724cm^-1,963cm^-1,1087cm^-1,1367cm^-1,1471cm^-1,1597cm^-1,1723cm^-1, but lower at828cm^-1,1002cm^-1,1032cm^-1and1128cm^-1; comparing with normal supernatant of homogenized tissues, the gastriccancer supernatant has higher Raman intensities at623cm^-1,656cm^-1,724cm^-1,963cm^-1,1087cm^-1,1367cm^-1,1471cm^-1,1597cm^-1and1723cm^-1, but lower at828cm^-1,1002cm^-1,1032cm^-1and1128cm^-1. The results indicate that there are highercontent of nucleic acides and lower content of most of proteins, carbohydrate andlipids in carcinoma tissues. The partial Least-squares based on the artificial neural network（PLS-ANN） isused to classify the spectra of normal and malignant tissues, homogenate andsupernatant of homogenized tissues. The sensitivity, specificity and accuracy oftissues are87%,81%and84%, respectively. The result of tissues is fine, but theresults of homogenate and supernatant of homogenized tissues are bad, the mainreason is the number of the examples of homogenate and supernatant of homogenizedare to small. And we need accumulate samples.This preliminary study demonstrates that it is possible to diagnose gastric cancerusing SERS technology in the future.