**Background:** The COVID-19 pandemic has highlighted limitations of current diagnostic methods, including the discomfort of nasopharyngeal swabs, delays in RT-PCR results (1-2 days), and reduced testing uptake due to fear of quarantine. Electronic-nose (e-nose) technologies offer a noninvasive alternative by detecting volatile organic compounds (VOCs) in exhaled breath that are altered by SARS-CoV-2 infection. This review summarizes evidence for e-nose-based COVID-19 detection and the underlying metabolomic biomarkers.
**Methods:** The authors conducted a narrative review of the literature on e-nose devices tested for COVID-19 detection, COVID-19-specific VOC biomarkers identified via metabolomic methods (GC-MS, PTR-MS, FTIR, etc.), and the pathophysiological mechanisms linking SARS-CoV-2 infection to altered breath VOC profiles. They evaluated seven e-nose devices: Aeonose (8 MOS sensors), NaNose (8 gold nanoparticle sensors), an experimental device (3 MEMS + 1 MOS), PEN3 eNose (10 MOS), EOS-AROMA (4 MOS), GeNose C19 (10 MOS), and Cyranose 320 (32 carbon black polymer composite sensors).
**Key Results:** Performance varied widely by device. The Cyranose 320 achieved 100% sensitivity and specificity (AUC 1.00) for post-COVID-19 syndrome detection in a pilot study of 123-132 patients. The GeNose C19 showed 88-95% accuracy (AUC 0.86-0.95) across two studies totaling over 1,000 samples. The NaNose achieved AUC 0.76 for COVID-19 vs. controls and AUC 0.95 for COVID-19 vs. other lung infections. The Aeonose had AUC 0.74 with negative predictive value 0.92. The PEN3 eNose performed poorly (AUC 0.58-0.63) with 57% false positives. The experimental 4-sensor device achieved AUC 0.79-0.89. The EOS-AROMA (4 MOS) had AUC 0.81 for respiratory failure vs. controls.
Metabolomic studies identified multiple quantitative VOC biomarkers consistently elevated in COVID-19 patients: aldehydes (heptanal, octanal, nonanal, ethanal, propanal), ketones (acetone, butanone), alcohols (ethanol, propanol), and alkanes. Heptanal was identified as the most consistent key biomarker. The FDA-approved InspectIR COVID-19 Breathalyzer (a portable GC-MS device detecting five undisclosed VOCs) showed 91.2% sensitivity and 99.3% specificity in a 2,409-person validation study.
**Clinical Implications:** E-nose technologies offer rapid (seconds to minutes), noninvasive, point-of-care screening that could complement or replace PCR in certain settings. The authors argue that screening effectiveness depends more on testing frequency and speed than on analytical sensitivity, making e-nose devices well-suited for frequent population screening. However, most tested devices were not specifically designed for COVID-19 VOC detection, and sensor drift remains a challenge for long-term monitoring. Only the InspectIR COVID-19 Breathalyzer has received FDA emergency use authorization (April 2022). The SpiroNose was approved in the Netherlands (February 2021). The authors recommend developing COVID-19-specific sensor arrays based on known VOC biomarkers and creating standardized reference databases to improve diagnostic accuracy.