Distinction of gastric cancer tissue based on surface-enhanced Raman spectroscopy

Abstract

ABSTRACT Gastric cancer is one of the most common malignant tu mors with high recurrence rate and mortality rate in China. This study aimed to evaluate the diagnostic capability of Surface-enhanced Raman spectroscopy (SERS) based on gold colloids for distinguishing gastri c tissues. Gold colloids were directly mixed with the supernatant of homogenized tissues to heighten the Raman signal of various biomolecule. A total of 56 samples were collected from normal (30) and cancer (26). Raman spectra were obtained with a 785nm excitation in the range of 600-1800 cm -1 . Significant spectral differences in SERS mainly belong to nucleic acid, proteins and lipids, particularly in the range of 653, 726, 828, 963, 1004, 1032, 1088, 1130, 1243, 1369, 1474, 1596, 1723 cm -1 . PCA-LDA algorithms with leave-one-patient -out cross validation yielded diagnostic sensitivities of 90% (27/30), specificities of 88.5% (23/ 26), and accuracy of 89.3% (50/56), for classification of normal and cancer tissues. The receiver operating characteristic (ROC) surface is 0.917, illustrating the diagnostic utility of SERS together with PCA-LDA to identify gastric cancer from normal tissue. This work demonstrated the SERS techniques can be useful for gastric cancer detection, and it is also a potential technique for accurately identifying cancerous tumor, which is of considerable clinical importance to real-time diagnosis. Keywords: spectroscopy; gastric cancer; Surface-enhanced Raman spectroscopy (SERS); gold colloids; principal components analysis (PCA); linear discriminant analysis (LDA)