Scientists have invented an ‘electronic nose’ that can identify COVID-19 infection through ‘smell’. The team, based at the Weizmann Institute of Science, Israel, began developing the innovative device at the start of the pandemic. Now, the technology is up to 94% accurate at detecting the presence of COVID-19 infection.
Researchers have developed an electronic nose that is effective at 'smelling' COVID-19. Image Credit: Andrii Vodolazhskyi/Shutterstock.com
The device could be fundamental to downgrading the pandemic by providing a simple, non-invasive, and portable method of diagnosing those infected with the SARS-CoV-2 virus. Additionally, there is potential for the device to be developed further so that it can diagnose other illnesses.
Diseases Give Off Distinct Smells
While the human nose may not be strong enough to know it, diseases have odors. For decades, dogs have been of interest in diagnostic research. They have successfully been trained to identify a variety of cancers, such as breast, bladder, and skin. More recently, they were trained to detect cancer by smelling a person’s breath.
The key to a disease’s smell is in the changes it makes to the body’s metabolic processes, resulting in different types and levels of metabolites being produced. In the case of COVID-19, virus-infected cells have been observed to produce volatile organic compounds (VOCs), which have a scent.
In an article published in 2020 in Nature, the ability of canines to smell COVID-19 was discussed. At that time, the evidence was promising, suggesting that dogs could smell a person infected with COVID-19 with high levels of accuracy. However, much of the data had yet to be peer-reviewed and there was the uncertainty of how this may be scaled up.
Now, scientists at the Weizmann Institute of Science have used this as inspiration to create electronic noses (eNoses) based on the animal olfactory system. While training and deploying dogs to key locations to smell out people who may be infected with COVID-19 seems costly and unlikely to work on a large scale, the deployment of eNoses is a simple and scalable strategy.
The team 3D-printed its eNoses and incorporated an array of sensors, each developed to efficiently detect a specific chemical range. The readout of their multi-sensor pattern can learn how to identify target patterns associated with specific viruses, bacterial infections, or other diseases. The new technology comprising a gas unit, sensor array, and a 3D-printed sampling valve has been successfully trained via practical testing to accurately identify COVID-19 as well as other VOCs within participants’ noses.
Searching for the Molecular Identity of COVID-19
The Israeli team used an AIRSENSE PEN3 eNose with thermo-regulated metal oxide sensors to allow it to ‘smell’ COVID-19. Then, they developed an additional one-way flow valve, which they 3D printed to prevent unintentional backflow, a problem that could lead to spreading the COVID-19 virus between users. Using CAD software, the scientists designed their sampling device so that it can fit flush against the nostril, where the air is pulled down towards the sensors.
To test the device, the eNose sampled the airflow from multiple patients attending Tel Aviv’s MADA testing clinic prior to taking a PCR test. The data was then run through software and trained to identify the differences in patterns between samples that came from participants who tested positive for COVID-19, and those who tested negative. After training it using a deep learning algorithm, the data showed that the eNose had achieved an average ‘true positive’ rate of 66.7% and an average ‘false positive’ result of 57%.
Currently, the team is yet to identify the exact molecular identity of COVID-19. Once found, the high rate of false positives may reduce to a level low enough to make the eNose a reliable diagnostic tool.
eNoses as a Future Diagnostic Tool
With more time in research and development, the eNose could improve sufficiently in accuracy to allow it to be reliably used to diagnose infection with COVID-19. It may provide an alternative, cheap, and rapid form of diagnosis that might help to manage the global pandemic. It has been suggested that this kind of testing could be useful at airports and live events to give an instant reading, highlighting those who may be infected and at risk of passing the virus onto many people.
With testing like this, it may help governments reduce restrictions that much of the population has lived under for more than one year. The eNose may be developed to diagnose infectious or non-infectious diseases.
References and Further Reading
Else, H., 2020. Can dogs smell COVID? Here’s what the science says. Nature, 587(7835), pp.530-531. https://www.nature.com/articles/d41586-020-03149-9
Snitz, K., Andelman-Gur, M., Pinchover, L., Weissgross, R., Weissbrod, A., Mishor, E., Zoller, R., Linetsky, V., Medhanie, A., Shushan, S., Jaffe, E. and Sobel, N., 2021. Proof of concept for real-time detection of SARS CoV-2 infection with an electronic nose. PLOS ONE, 16(6), p.e0252121. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0252121
3D printing-enhanced ‘enose’ trained to ‘sniff-out’ covid-19 in just 80 seconds. Paul Hanaphy. 3D Printing Industry. Available at: https://3dprintingindustry.com/news/3d-printing-enhanced-enose-trained-to-sniff-out-covid-19-in-just-80-seconds-191080/