Answers are grounded in this paper and cite their sources.
Exhaled volatile organic compounds (VOCs) hold great promise as non-invasive biomarkers for disease diagnosis. However, conventional exhaled VOCs sampling and analytical methods suffer from significant limitations in background noise, sample stability, sensitivity, and resolution. Moreover, shared VOCs signatures across different diseases poses challenges to accurate diagnosis within single-disease models. These limitations hinder accurate profiling of the full spectrum of trace VOCs and constrain their large-scale clinical application. To address this technological gap, we developed a comprehensive exhaled VOCs analysis platform integrating a self-developed BreathScope sampler and a GC × GC-TOF MS/FID detection system, complemented by optimized sampling strategies and standardized workflows. Compared with conventional methods, BreathScope enhanced both the diversity and concentration of captured VOCs by precisely targeting end-tidal breath, effectively reducing interference from exogenous compounds. The analytical system successfully identified hundreds of trace VOCs, demonstrating high quantitative accuracy (R² > 0.97), precision (RSD < 10%), with a detection and quantitation limit at the ng/L level. Additionally, we identified the optimal sample volume and demonstrated the impact of confounding factors on VOC profiles, highlighting the necessity for standardized sampling protocols. Utilizing exhaled VOCs data from 509 subjects, we constructed a multi-class random forest model for risk assessment of colorectal, gastric, and brain cancers, achieving AUROC values of 0.98–0.99, with sensitivity and specificity exceeding 0.95. Altogether, our platform integrates high-fidelity breath collection, trace VOCs quantification, and AI-driven disease prediction, thereby enhancing the utility and standardization of breath biopsy for early disease screening and offering a reproducible technological framework for non-invasive precision medicine.
No takes yet. Share an insight, caveat, or question.
Li et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: