Comparison of SERS spectral data sets of blood serum samples of hypopharyngeal cancer using silver and gold nanoparticles as substrates

BACKGROUND: Early detection of hypopharyngeal cancer is crucial for a good prognosis and survival, as it is mostly diagnosed at an advanced stage due to subtle early symptoms. Blood serum contains valuable diagnostic biomarkers for disease screening as it contains metabolites from various tissues and organs in the body and represents the biochemical changes occurring due to disease development. RESEARCH DESIGN AND METHODS: Surface-enhanced Raman spectroscopy (SERS) was applied to blood serum from healthy individuals and patients at different stages of hypopharyngeal cancer. Both silver (AgNPs) and gold nanoparticles (AuNPs) were used as SERS substrates to compare their SERS features. Principal component analysis and linear discriminant analysis (PCA-LDA), along with support vector machine (SVM) were performed to analyze and differentiate the features of both substrates. RESULTS: The analysis revealed complementary SERS spectral features associated to specific nanoparticles as a substrate for cancer diagnostics and metabolic profiling. The SVM classifier achieved an overall accuracy of >90%, with a macro-averaged AUC of 0.98. The class-wise performance metrics (precision, recall, and F1-score) approached unity, which confirms the robust discrimination across all diagnostic groups. CONCLUSIONS: The comparison of SERS spectra of both substrates identified that both substrates differentiate the healthy from cancerous ones while providing different and complementary spectral features.