Applying the electronic nose for pre-operative SARS-CoV-2 screening.

Area of Science:

• Biomedical Engineering
• Infectious Diseases
• Computational Biology

Background:

• Standard COVID-19 diagnosis relies on RT-PCR, which can yield false negatives.
• A need exists for improved, non-invasive diagnostic methods for SARS-CoV-2.
• Volatile organic compounds (VOCs) in breath may indicate infection status.

Purpose of the Study:

• To evaluate an electronic nose (Aeonose) for differentiating COVID-19 positive and negative individuals.
• To assess the potential of machine learning classifiers trained on VOC patterns for COVID-19 detection.

Main Methods:

• Breath samples were collected from participants undergoing RT-PCR testing.
• An electronic nose (Aeonose) analyzed VOCs in exhaled breath over five minutes.
• Machine learning algorithms processed sensor data to generate an infection probability score.

Main Results:

• The study included 219 participants, with 57 testing positive for COVID-19.
• The Aeonose achieved a sensitivity of 0.86 and a negative predictive value (NPV) of 0.92.
• Incorporating clinical data improved the NPV to 0.96.

Conclusions:

• The Aeonose demonstrates capability in distinguishing COVID-19 status via VOC analysis.
• This electronic nose shows potential as a non-invasive, low-cost triage tool for excluding SARS-CoV-2 infection, particularly in surgical patients.