Minimally Invasive Raman Spectroscopy for Detection of Lymph Node Metastasis of Oral Squamous Cell Carcinoma

Lymph node metastasis (LNM) is a critical prognostic factor in oral squamous cell carcinoma (OSCC), yet its preoperative assessment remains challenging. This study evaluated serum and salivary Raman spectroscopy (RS) for LNM detection in OSCC. Serum and saliva samples from 36 OSCC patients [16 node-positive (N+) and 20 node-negative (N0)] and 15 healthy controls were analysed using RS. Classification models were developed using RS data of individual biofluids and with fusion strategies. The serum model identified 87.5% of N+ and 85% of N0 cases, while saliva showed comparable performance for N+ but lower accuracy for N0. Low-level fusion improved N+ classification. Mid-level fusion achieved the best overall performance, correctly identifying 100% of N+ and 85% of N0 cases. Spectral analysis revealed alterations associated with proteins, β-carotenoids, deoxyhaemoglobin and lipids across the study groups. These findings demonstrate the potential of biofluid-based RS for minimally invasive LNM assessment in OSCC.