Raman Spectroscopy of Biofluids: Fusion Strategies for Detecting Oral Potentially Malignant Disorders and Oral...

Area of Science:

• Biomedical Spectroscopy
• Diagnostic Technologies
• Cancer Research

Background:

• Early detection of oral cancers is crucial for reducing mortality.
• Differentiating oral potentially malignant disorders (OPMDs) from oral cancers is a significant clinical challenge.
• Minimally invasive screening tools are needed for effective oral cancer diagnosis.

Purpose of the Study:

• To evaluate the efficacy of serum and salivary Raman spectroscopy (RS) for detecting OPMDs and oral cancers.
• To compare the diagnostic performance of serum RS versus saliva RS.
• To investigate the impact of data fusion techniques on improving classification accuracy.

Main Methods:

• Serum and saliva samples were collected from controls, tobacco users, leukoplakia, oral submucous fibrosis, and oral squamous cell carcinoma patients.
• Raman spectra were acquired and analyzed using multivariate statistical methods.
• Serum and saliva data were analyzed individually and in combination using low-level and mid-level data fusion.

Main Results:

• Serum RS achieved high individual classification accuracies for all groups (up to 100% for leukoplakia).
• Saliva RS showed limitations in differentiating between certain conditions.
• Mid-level data fusion of serum and saliva RS significantly enhanced classification accuracy, reaching 100% for controls, leukoplakia, and oral submucous fibrosis.

Conclusions:

• Serum and salivary Raman spectroscopy hold potential as non-invasive screening tools for OPMDs and oral cancers.
• Data fusion techniques, particularly mid-level fusion, offer a powerful approach to improve diagnostic performance.
• Further research is warranted to validate these spectroscopic methods for clinical application in oral cancer screening.