Agnihotri, Aradhya2024-11-252024-11-252024-11-25http://hdl.handle.net/10393/49887https://doi.org/10.20381/ruor-30712This thesis focuses on the development of a Surface-Enhanced Raman Spectroscopy (SERS) assay for detecting biofilms in ocular health. Biofilms are complex communities of microorganisms encased in an extracellular polymeric substance (EPS) that present significant challenges for clinical diagnostics due to their resistance to antimicrobial treatments and their evasion of traditional detection techniques. By amplifying Raman signals through metal nanoparticles, SERS offers a promising solution for overcoming these challenges in detecting biofilm-associated infections in tissues such as the cornea. Silver nanoparticles (AgNPs) were synthesized and optimized to interact with biofilms using a factorial design approach. Optimal conditions of 1.75x nanoparticle concentration and a 24-hour interaction time between the sample and the nanoparticles were identified. This combination produced the most robust signal, particularly for detecting pyocyanin, a key virulence factor in Pseudomonas aeruginosa (PA01) biofilms. The novelty of this work is the application of this optimized protocol to pig cornea samples, which closely mimic human corneal composition. The SERS spectra from the corneal samples were consistent with earlier validation results we achieved from the collagen samples, with some expected differences due to the distinct structural composition of the cornea. Machine Learning (ML) techniques were integrated into the SERS data analysis to enhance the diagnostic power of the assay. By combining unsupervised methods like K-means clustering with supervised techniques such as Support Vector Machines (SVM), the model effectively differentiated biofilm-present from biofilm-absent samples, achieving an overall accuracy of 92.5%, specificity of 100%, and sensitivity of 80%. This integration of SERS and ML presents a significant advancement for real-time, minimally invasive diagnostics in ocular health. This research establishes a strong foundation for using SERS in biofilm detection within ocular health, demonstrating the potential of an optimized SERS assay combined with ML techniques to improve early detection and treatment of ocular infections such as bacterial keratitis. However, challenges remain in managing biofilm composition variability and addressing spectral overlap with corneal tissue. Future work will focus on refining diagnostic algorithms to enhance accuracy and sensitivity, expanding the assay's scope to detect a broader range of bacterial infections, and advancing toward human clinical trials.enAttribution-NonCommercial 4.0 Internationalhttp://creativecommons.org/licenses/by-nc/4.0/Surface-Enhanced Raman Spectroscopy (SERS)Silver NanoparticlesBiofilm DetectionCorneal InfectionsMachine LearningDiagnosticsBiomedical ApplicationsClusteringClassificationThesis