JMIR Publications

ABSTRACT

Early diagnosis of oral squamous cell carcinoma (OSCC) is critical for improving outcomes, yet biopsy is invasive and imaging has limited sensitivity. Exhaled breath analysis of volatile organic compounds (VOCs) offers a non invasive alternative for early screening. This systematic review searched PubMed, Embase, and other databases, identifying 16 studies. We examined four aspects: sampling, detection technologies, data analysis, and biomarkers. Mixed VOC patterns showed better diagnostic performance than single biomarkers. Electronic noses demonstrated high sensitivity and specificity. Proton Transfer Reaction-Time-of-Flight Mass Spectrometry (PTR-TOF-MS) offers real-time online monitoring, requires no sample pretreatment, achieves pptv-level sensitivity, and provides high mass resolution. GC IMS and the combination of TD GC MS with GC IMS performed well across different sample types. Machine learning algorithms (Random Forest, SVM, ANN) performed robustly on small high dimensional datasets, with cross validation and permutation tests effectively mitigating overfitting. However, sampling standardization, mechanistic validation, and algorithm interpretability remain major barriers to clinical translation. Future work should establish international SOPs, elucidate VOC biological pathways, and develop interpretable AI models to enable clinical translation for OSCC screening.