Study On The Diabetes Detection By Surface-enhanced Raman Spectroscopy Of Urine

Raman spectroscopy is a kind of testing method, using a non-elastic scattering light generated during mutual action of photon and material molecule. Raman spectroscopy is non-invasive and is highly accurate for analyzing structure of substances, so it has been applied widely in many fields:such as biomedicine, material science, environment, chemistry and so on. As the discovery and development of Surface-Enhanced Raman spectroscopy (SERS), Raman spectroscopy has been gradually used to detect the disease tissue and body fluid.This paper introduces the basic theory and the development process of Raman spectroscopy, elaborates the application and study progress of Raman spectroscopy in biomedicine field. This paper focuses on the analysis of Surface-Enhanced Raman spectroscopy detection of diabetic and normal urine. The number and location of the Raman scattering peaks, principal component analysis and discriminant analysis were used.Analyzing the Raman spectrum data we could obtain several conclusions. Whether the urine in the 1119～1122 cm-1,1414～1417 cm-1 two band appearing peaks at the same time is used as a criterion, we can distinguish the diabetes from normal people, the accuracy of diabetes diagnosis is 90%. Define Fi=Ii/It,Ii, is the highest point of the Raman spectrum peak intensity value,It is the corresponding Raman spectrum lowest intensity value), Using 1004cm-1,1049cm-1,1084cm-1,1120cm-1,1414 cm-1,1553cm-1,1577cm-1 SERS peaks'Fi as a variable, we apply PCA-LDA by SPSS software to analyze. Using principal component analysis, the accuracy of the diagnosis is 90%; while the error probability of healthy people is 0 and the accuracy of the diagnosis is 85% if using discriminant analysis. These methods can provide some references for the clinical diagnosis of diabetes.Through consulting extensive literature, this paper presents several Raman scattering peaks initial ownership, we think the peak between the 1121cm-1 and 1415cm-1 maybe caused by the albumin, the 1120cm-1 peak maybe come from Urea, while the peak at 1416cm-1 maybe produced by the creatinine. Ownership of other peaks needs further research and analysis.