(683b) Noninvasive Detection of COVID-19 By Analysis of Exhaled Breath

Background: The World Health Organization (WHO) declared COVID-19 outbreak caused by SARS-Cov-2 a pandemic in March 2020. The ongoing crisis has caused over 701 million infections and 6.96 million deaths worldwide. The global pandemic has put a spotlight on the urgent need for an accurate, rapid and noninvasive test for COVID-19 because of the highly contagious nature and health effects of the disease.

Objective: The objective of this work is to develop a technique using silica cartridges for selective capture carbonyl compounds in exhaled breath for COVID-19 detection and using gas chromatography-mass spectrometry (GC-MS) for the identification.

Methods: We collected exhaled breath from COVID-19 positive and negative subjects confirmed by reverse transcriptase polymerase chain reaction between April 2022 and September 2023. A total of 237 subjects were enrolled in this study. Of these, 142 were COVID-19 positive and 95 were negative. A machine learning workflow was implemented with the captured carbonyl compounds as input features for diagnosis.

Results: The proposed approach classified diagnosis with an accuracy of 95.4%, sensitivity of 96.5%, specificity of 93.7%, and AUROC=0.974. 10 features including 3 carbonyl compounds and 7 ratios show significant differences between COVID-19 positive and negative groups.

Conclusion: The analysis of carbonyl compounds in exhaled breath has great potential for non-invasive COVID-19 diagnosis using silica cartridge approach.