Raman spectroscopy for diagnostic tissue assessment

Current gold standard treatments for diseases such as post-traumatic osteoarthritis (PTOA) and cirrhosis are qualitative and invasive. The limitations of these surgeries were addressed by developing an objective and reliable detection of these diseases. The use of Raman spectroscopy as a diagnostic tool in a clinical setting could enable quantitative analysis of connective tissue and eventually improve surgical outcomes. This technology is based on statistical multivariate modelling of the interaction between tissue and Raman scattered light. In this thesis, two studies are presented: the quantitative analysis of ligament viscoelastic properties and the assessment of liver pathologies and anatomical features for assessing liver transplants. The first aim of this thesis is to investigate whether functional properties of collateral ligaments from rabbit joints can be quantitatively evaluated using Raman spectroscopy. Previously it has been shown that near-infrared (NIR) spectroscopy can enable the quantitative assessment of articular cartilage and has been proposed as a diagnostic tool for arthroscopy. While there is an abundance of evidence to support NIR spectroscopy’s clinical use in quantitative arthroscopy due to its capability in characterizing articular cartilage, there is lacking investigations in Raman spectroscopy’s application in evaluating knee joint tissue particularly ligaments. The results of this investigation demonstrate that Raman spectroscopy could accurately estimate the viscoelastic properties of ligaments using partial least squares (PLS) regression with a relatively low root-mean-square error (RMSE). The measurement of the viscous damping properties of ligaments provides clinicians with a reliable indicator for structural integrity. A study of the viscoelastic properties of ligaments along with its application into arthroscopy will vastly improve diagnostic outcomes, patient quality of life, and our understanding of the pathophysiological changes that take place in diseased joints. The second study conducted as part of this thesis involved the collection, preparation, and spectral analysis of human liver tissue. While biopsy is considered the gold standard approach to the treatment and diagnosis of liver pathology, it does not provide quantifiable and objective molecular data. As Raman spectroscopy has shown increasing use as an analytical tool in the medical field, it was hypothesised that this modality could detect the structural and compositional changes associated with liver pathology. These pathological changes could involve altered metabolism, inflammation, fibrosis, or fat accumulation (steatosis). Using traditional multivariate methods such as support vector machine (SVM) and performance metrics such as accuracy, recall, precision, and f1-score, pathologies associated with the interlobular bile ducts of hepatocytes could be detected with Raman spectroscopy. Potential modifications to the composition could include alterations in lipid metabolism, protein conformation, bilirubin, bile acid, and other biochemical components that are indicative of pathological processes. In conclusion, Raman spectroscopy has potential to identify the molecular signatures of processes such as fibrosis, steatosis, and inflammation and by extension, detect structural and compositional changes associated with pathologies of the interlobular bile duct. This technology shows great potential particularly in the fields of orthopedics and hepatology. Because the pathology of the tissue alters its optical and scattering properties, Raman spectroscopy will enable the rapid and non-destructive assessment of the molecular composition, structural integrity, and overall health of the tissue.