Skip to main content
Premium Trial:

Request an Annual Quote

Finland-Based Deep Sensing Algorithms Works to Commercialize COVID-19 Breath Test After CE Marking


NEW YORK – Finnish diagnostic company Deep Sensing Algorithms (DSA), like many other diagnostic companies in the COVID-19 testing space, didn't exist before the pandemic. 

According to Pekka Rissanen, DSA's CEO, the company grew out of a "coffee table discussion" in the spring of 2020, where he and his cofounders questioned what they could do to help people during the pandemic. The team "had a hunch" COVID-19 would be around longer than anyone expected, and by June of that year the company, a "purpose-built venture" intended to develop and commercialize a breathalyzer test for COVID-19, was born.

Rissanen's colleagues in the company had previous experience using nanosensors to detect volatile organic compounds, and it took only a few months to develop five prototypes of what would eventually become the DSA BreathPass, Rissanen said. Those prototypes were distributed to sites across the globe to "get the devices in front of people" and begin validating the product, he said.

By the end of 2020, the company had developed additional prototypes, 100 of which were distributed to 40 different countries to conduct preclinical trials of the test. The actual finished product began undergoing preclinical trials in Spain and the Netherlands almost exactly a year ago, and data from a clinical trial conducted in Spain in October was used for the firm's CE mark application. The test received CE marking on Jan. 14, and since then the company has been further validating and verifying the test before mass deployment, Rissanen said. 

So far, DSA has received orders from multiple countries both inside and outside of Europe, including New Zealand, Venezuela, Mexico, Turkey, Finland, Germany, and Spain, Rissanen said. The firm is also "in talks with" Health Canada and Australia's Therapeutic Goods Administration for regulatory approval in those countries, he added. 

From discussions with the US Food and Drug Administration, DSA has also discovered steps it will need to take to achieve regulatory compliance to offer its test in the US, including a specific compliance measure that the firm hasn't implemented yet, but which it expects to take in six to nine months, Rissanen said. DSA intends to seek Emergency Use Authorization from the agency for the test in final pursuit of 510(k) clearance down the line.

Commercialization in the US is one of the ultimate goals for the company, but it's targeting outside the US first because of easier regulatory processes elsewhere, according to Rissanen. 

The test itself combines off-the-shelf hardware and in-house software to detect a patient's metabolic response to COVID-19. Using a list of target biomarkers found in the breath of patients with COVID-19, the test can determine a positive or negative result. DSA BreathPass is "not so interested in one chemical compound," Rissanen said, but rather a "particular pattern" in the breath that includes between 14 and 20 gases related to COVID-19. 

Each person has a "breath print," comparable to a fingerprint, and the BreathPass uses its nine sensors and a proprietary algorithm to take that breath print and determine whether it indicates the presence of COVID-19, Rissanen said. The machine-learning algorithm was built on samples of both healthy patients and those with COVID-19 to build a profile of gases to compare to an individual patient's sample. 

All the test requires is the device and its associated app, which tells patients exactly how to use the test, Rissanen noted. A user exhales once into the device and the sensors analyze that breath for 45 seconds, then send the analysis through the cloud to the algorithm, which takes three seconds to return a final result to the app, he said. The entire cycle takes approximately two minutes. The device includes an air filter to keep remnants of other molecules out and a one-way valve to prevent accidental inhalation.

According to the data the firm has collected from its validation studies, BreathPass has a sensitivity of around 80 percent for asymptomatic patients. Although that is lower than the stated performance of many antigen tests currently on the market, Rissanen said that many of those tests have shown worse performance in the real world than their companies have claimed. He cited a study published in March 2021 in the Cochrane Database of Systematic Reviews that found antigen tests were 58 percent sensitive in asymptomatic patients, noting that DSA BreathPass would outperform those tests and that "good enough is good enough." 

Asymptomatic patients in the early stage of their infections are the key population for the DSA BreathPass and a group with which some antigen tests have appeared to struggle. The test is geared toward the early stage of infection where the metabolic signals it's measuring are strongest, Rissanen said. Once symptoms present, there are additional volatile organic compounds and signals that could throw a wrench into what the test is detecting and make it harder to read the specific pattern the algorithm is looking for, he said. 

According to Rissanen, the test's performance isn't impacted by the known SARS-CoV-2 variants, and he noted that BreathPass doesn't detect which variant is present — just whether the virus itself is. 

He also said that the company isn't distributing the device, but rather working with distributors to get it in the hands of consumers. The device will cost the same as an "upscale mobile phone" and has a lifetime of three to five years. An individual test, meantime, will cost about the same as a home test, he added.

Rissanen said the test could conceivably be used in a variety of settings, including at work, school, and places where services can't be offered remotely, like dentists' offices. However, the sensors within the device are particularly sensitive to alcohol and work best in places where there's some control over the airflow to ensure a relatively clean environment, he said — deploying the test at an airport wouldn't be suitable, for example. 

Beyond COVID-19, the firm has long-term plans to develop a product for cancer screening, although Rissanen declined to specify what type of cancer was on the table, saying only that there was "a lot of interest on the idea level."

Although it has made significant progress since it began, DSA isn't the only company interested in breath testing for COVID-19 and cancer. Singapore-based Breathonix has developed a mass spectrometry-based breath test for SARS-CoV-2 that has obtained regulatory approval in Singapore and Malaysia, as well as CE marking in November. The firm is currently working toward regulatory approvals in the Middle East. 

Nanotechnology firm Imec, meantime, has developed a silicon chip-based breath test for SARS-CoV-2, while Owlstone Medical has long been identifying and validating breath-based biomarkers for a range of diseases, including lung cancer, to be used in developing diagnostic tests. And Belgium's MiDiagnostics is in the process of developing a rapid breath test using qPCR that is expected to launch this year.