Research On The Surface Enhanced Raman Spectroscopy Of Saliva For The Diagnosis Of Lung Cancer

Lung cancer is the leading cause of morbidity and mortality worldwide, the low survival rate of lung cancer is the difficulty to be detected at an early stage. Mostly used screening techniques such as sputum detection, bronchoscopy, and low-dose spiral CT scans all have some defects like risk of over diagnosis, lung collapse, and false positives. As such, 80% of the patients have been diagnosed malignant at the first time they go to see the doctor. Surface enhanced Raman spectroscopy (SERS) has the advantages of high sensitivity, high resolution and non-invasiveness, accompanying the increased enhancement effect provided by highly ordered SERS substrate, it has been widely used in the disease diagnosis. Saliva is the ultrafiltration of blood and has approximately the same components and has been used for the diagnosis of many disease. In this thesis, saliva of 21 lung cancer patients and 22 normal people were tested using SERS. The statistical analysis of PCA-LDA and the assignment analysis of Raman peaks were applied on the two groups of preprocessed SERS. In the PCA-LDA, an accuracy of 86.05%,a sensitivity of 94.12% and a specificity of 80.77% were obtained. Referring to peak assignments of tissue, body fluid, and biological standards in literature, Raman peaks were assigned. SERS peaks at the position of 523 cm^-1, 623 cm^-1, 696 cm^-1 and 905 cm^-1, intensity decreased 27.12%,29.58%,33.66% and 25.74% separately from normal people to lung cancer patients. They were assigned to S-S stretching, C-S stretching, C-C twisting and the ring bending vibration, which were devoted by proteins and nucleic acid bases (A, T, C, G). The results indicated that the content of proteins and neucleic acids of lung cancer saliva is lower than that of normal people. The results are in agreement with the analysis on the in vitro lung cancer cells. At the position of 932 cm^-1, the Raman peak of lung cancer increased a little. Most literature assigned it to C-C twisting vibration of proline and valine. This indicate an increase of certain proteins or strutural changes of some proteins. In comparison with the clinical medical imaging technology based on tissue morphology, the diagnosis method stated in this thesis is an analytical tool that can detect the variation of contents and conformation on the molecular and cellular level. With the features of trace detectability, early stages, short detection time, easy to use, non-invasiveness, easy to screening and dynamic monitoring, this method is a development direction of medical diagnosis technology and method.