A newly developed electronic ‘nose’ could identify COVID-19-infection

By James DuckerReviewed by Emily Henderson, B.Sc.Jun 2 2021

Developing rapid diagnostic tools is key to dealing with the COVID-19 pandemic, and a new electronic-based ‘nose’ detector could help identify cases in 80 seconds.

Testing of the "electronic nose". Image Credit: Kobi Snitz

A faster, cheaper, and better way for COVID-19 testing

The COVID-19 pandemic has generated a surge of interest in developing rapid, cost-effective, and readily deployable detection of infections. Such measures would allow for better management of infections and better tracing capabilities.

However, despite increased research efforts attempting to increase the speed and efficacy of detecting COVID-19, the most commonly used method remains PCR testing. Although this provides accurate results, the duration, training, and setup required makes this technique potentially difficult to readily deploy on a large scale. As such, governments have focused testing efforts at key focal points to concentrate patients such as drive-throughs or large-scale testing clinics.

In a new effort led by researchers at the Weizmann Institute in Rehovot, Israel, a team led by Professor Noam Sobel, the Head of the Department of Neurobiology, is now developing an electronic-based tool to detect COVID-19 infections using a faster, cheaper, and more efficient method.

Electronic detection of infectious diseases is a long-standing concept, with research labs attempting to adapt the method to COVID-19 over the past year. The technique relies on the characterization and detection of metabolites by devices that can identify the ‘smell’ of metabolites produced by the metabolic processes of infections.

Nevertheless, COVID-19 has proven difficult to detect as it is a disease with metabolites that are particularly difficult to characterize. In response, most research has focused on lab-based methods using samples already known to be infectious, a concept that Prof. Sobel wants to overturn.

The Weizmann Institute team transitioned from lab-based to field-based testing, adjusting their methods to better adapt to more deployable strategies. The current device in testing is a plastic 3D printed tool containing a one-way valve in which patients can breathe through, similar in appearance and size to a breathalyzer, connected to a laptop. Researchers then set up at a nearby testing ‘drive-thru’ facility set up by the Israelian Red Cross, collecting data from individuals being tested immediately as they arrive.

This method can provide results within 80 seconds of testing, providing rapid and readily available infection data much faster than PCR tests currently in place with a much more reduced setup needed.

Difficulties in teaching a ‘nose’ to smell COVID-19

Data is currently being collected, analyzed, and sorted from patients which is then gathered into a COVID-19 learning program for the electronic ‘nose’ to learn to detect infected individuals. This teaching is currently in progress, and results on the effectiveness of detection may take some time.

Indeed, teaching this new device has proven difficult as the characteristic metabolite profile, or ‘smell’, of COVID-19, has yet to be found. Typical electronic detection works with a specific endpoint, a known disease profile from which the program can learn. However, the COVID-19 profile has yet to be fully characterized. This increases the time needed for the learning process as it lacks the successful endpoint that typical machine learning requires.

Nevertheless, prolonged teaching of COVID-19 detection is important to optimize identification rates, providing more accurate tests. Moreover, the field-based method of the research team provides faster, and more data than lab-based research, aiming to accelerate the learning process of this electronic ‘nose’.

Field-based methods will also provide larger datasets with key data on the individual variation of infection characteristics, such as infection levels, which in turn, will help testing facilities to better identify individuals or areas at risk.

Researchers are confident that this new tool will provide a new readily available testing method once the COVID-19 metabolite profile has been learned. Further tests remain to be conducted on the new variants emerging in different countries, yet this technique would significantly improve the speed, cost, and deployment of COVID-19 testing.