Raman Biosensors

Abstract

This PhD thesis focuses on improving the limit of detection (LOD) of Raman biosensors by using surface enhanced Raman scattering (SERS) and/or hollow core photonic crystal fibers (HC-PCF), in conjunction with statistical methods. Raman spectroscopy is a multivariate phenomenon that requires statistical analysis to identify the relationship between recorded spectra and the property of interest. The objective of this research is to improve the performance of Raman biosensors using SERS techniques and/or HC-PCF, by applying partial least squares (PLS) regression and principal component analysis (PCA). I began my research using Raman spectroscopy, PLS analysis and two different validation methods to monitor heparin, an important blood anti-coagulant, in serum at clinical levels. I achieved lower LOD of heparin in serum using the Test Set Validation (TSV) method. The PLS analysis allowed me to distinguish between weak Raman signals of heparin in serum and background noise. I then focused on using SERS to further improve the LOD of analytes, and accomplished simultaneous detection of GLU-GABA in serum at clinical levels using the SERS and PLS models. This work demonstrated the applicability of using SERS in conjunction with PLS to measure properties of samples in blood serum. I also used SERS with HC-PCF configuration to detect leukemia cells, one of the most recurrent types of pediatric cancers. This was achieved by applying PLS regression and PCA techniques. Improving LOD was the next objective, and I was able to achieve this by improving the PLS model to decrease errors and remove outliers or unnecessary variables. The results of the final optimized models were evaluated by comparing them with the results of previous models of Heparin and Leukemia cell detection in previous sections. Finally, as a clinical application of Raman biosensors, I applied the enhanced Raman technique to detect polycystic ovary syndrome (PCOS) disease, and to determine the role of chemerin in this disease. I used SERS in conjunction with PCA to differentiate between PCOS and non-PCOS patients. I also confirmed the role of chemerin in PCOS disease, measured the level of chemerin, a chemoattractant protein, in PCOS and non-PCOS patients using PLS, and further improved LOD with the PLS regression model, as proposed in previous section.