Coronavirus: Using odors to detect an infection

Rapid COVID-19 tests can be rather uncomfortable as samples are typically collected with a deep nasal or throat swab. Scientists now explore an alternative to rapid diagnostic tests based on a patient's breath.

In this MEDICA-tradefair.com interview, Dr. Sybelle Goedicke-Fritz explains how diseases can be detected through smell and reveals how electronic nose technology can be used for medical applications.

Dr. Goedicke-Fritz, you are working on an application that detects an infection with the coronavirus via odor. How does this technology work?

Dr. Sybelle Goedicke-Fritz: The ancient Greek physician Hippocrates discovered that diseases produce distinct odors in people, making them smell differently than healthy persons. We all emit so-called volatile organic compounds (VOCs). Infections can cause a change in these vapors. One cause might be the activation of the body's immune system in response to a new infection, or the bacteria causing the infection.

And these compounds are released through the lungs?

Goedicke-Fritz: That depends on the disease pattern. Respiratory infections obviously affect the lungs, but we are also studying sweat and saliva, for example. Lung cancer has also already been detected through exhaled air. However, you must always consider the disease location.

Are there any indications or evidence that this also works in detecting COVID-19?

Goedicke-Fritz: A study based in Hanover, Germany, used dogs to detect COVID-19 in patients based on sweat samples. We now want to achieve this using an electronic nose.

What device are you using in this setting?

Goedicke-Fritz: We use the Cyranose 320, which is used by the US military to detect the presence of explosives. The device comprises 32 individual polymer sensors, which are activated in varying degrees once they are exposed to VOCs. This creates distinct aroma-signature patterns. These are visualized as so-called "smellprints" in 3D graphical representation.

Is the device being used for medical applications?

Goedicke-Fritz: The device is not yet established as a medical device, though it is already being tested for lung cancer detection. We are currently testing it pertaining to amnionitis. However, we are still in a testing phase as this is essentially artificial intelligence that needs lots of data to train and learn.

Our objective is to teach the device to detect the smellprint for COVID-19.

Is your plan to use the device primarily to screen children for the disease?

Goedicke-Fritz: We work at the Department of General Pediatrics, which means we primarily deal with children. A key issue is that there are age limits as it relates to COVID-19 vaccines for children and transmission tends to occur in schools. What's more, rapid COVID tests are simply too uncomfortable for children, especially for little ones, which is why our efforts focus on this group.

Having said that, we also factor adults into this setting and collaborate with pulmonary critical care.

Your project received funding from the Else Kröner-Fresenius Foundation in November 2021. What are your next steps?

Goedicke-Fritz: We have already submitted our application for the study to the ethics committee. Our goal is to start the measurement process on children in January. All told, we want to include 250 patients to facilitate solid data collection. In the summer, but no later than the end of 2022, we aim to implement the software improvements. Project completion should be in two years.