CHICAGO, Ill. -- By Lee Han-soo/Korea Biomedical correspondent – Lunit is well-known among the Radiological Society of North America (RSNA) community for its top-notch artificial intelligence (AI) technology in the field of medical imaging.

The company was one of the first Korean AI medical imaging companies to attend the RSNA as an exhibitor in 2016 and has since unveiled several new products and a variety of clinical studies every year.

The company continued to boast its prowess during RSNA 2022, with the event organizers selecting Lunit’s 12 clinical studies on its various products – eight oral presentations and four posters. The figure accounted for the most abstracts released by a Korean medical device company during the RSNA 2022.

Notably, the company garnered much attention from the radiology society after releasing the first prospective study using medical AI imaging on Sunday.

Due to the company's active activities in the clinical field, Lunit's booth at RSNA 2022 was packed with participants making inquiries about using its device.

In light of such high interest, Korea Biomedical Review met with Lunit's Chief Medical Officer (CMO) Kim Ki-hwan at RSNA 2022 on Monday to get more in-depth knowledge of the company's progress during the convention and its plans.

Question: Despite only marking the second day at RSNA 2022, the company's booth seems crowded with visitors. I am curious about the reaction of RSNA 2022 participants to this year's Lunit booth and research presentations.

Answer: As we have been participating in RSNA for quite some time, radiologists participating in the convention, thankfully, have shown great interest in our company this year.

I believe that this is because the AI medical device industry is changing.

Only three to four years ago, medical device companies could receive attention by even the slightest mention of AI applications on their products.

However, with the level of participants' knowledge of AI increasing and the application of AI in the field prevalent, only a handful of companies that have achieved certain goals, such as product approval, have started to receive attention from radiologists.

So, while many radiologists came to our booth with vague curiosity about AI in the past, radiologists who visit our booth this year have a clearer understanding of the product, have heard of us from colleagues that use our product, or are interested in conducting joint research using our product.

Q: What are some of Lunit's expectations or goals from participating in RSNA 2022?

A: Unlike last year, we made a private room in the booth this time.

While we focused a little more on promoting our AI software in the past, this year, we are looking to hold a lot of meetings with radiologists and foreign medical device companies to expand our business.

From an academic point of view, in the past, we tried to focus on increasing the number of abstract presentations during the RSNA convention, as it was the only way for a small company to receive attention from radiologists.

However, this year, we are trying to present abstracts that have a profound impact on the industry. I think such high-level research will be our own differentiation point moving forward.

Also, in recent years, rather than Lunit-led research announcements, our customers or radiologists who use our products are conducting research on their own and presenting research results.

Therefore, while we continue to research and present clinical studies on new AI systems under development, we plan to minimize academic activities for already commercialized products.

Rather, we want our business partners or researchers to conduct research and publish the results using their own voices without any interference from us.

Q: On Sunday, the company presented the first prospective study using medical AI imaging, where a Swedish research team applied Lunit Insight MMG, a mammography image analysis solution, at actual medical sites. How valuable is the research?

A: I think the value of the prospective study result is equivalent to a pharmaceutical company having a successful phase 3 clinical trial for a new drug.

Now, if you think about what a high-quality study is, it should be a prospective study.

However, in the field of radiology, there are not many prospective studies.

Now, unlike the pharmaceutical or bio-clinical side, most of the studies that are mainly conducted on the radiology side look at how much the diagnostic performance differs from the data collected retrospectively.

We are well aware that there are some opinions among radiologists that a prospective study is not necessary.

However, we want to show how AI-based medical imaging can be used in the actual medical field.

As a result, a prospective study in a typical experimental environment was absolutely necessary as it could serve as the ultimate clinical evidence for using our system.

Even in the case of the recently presented study by Professor Fredrik Strand at the Karolinska Institutet, the study proved that our AI could reach the level of efficacy that radiologists require when they introduce the system to actual breast cancer screening.

While the diagnosis method differs by country, in Europe, two doctors have to read the image to make a diagnosis, but Professor Strand's research proved that one doctor plus our AI system could do the same job with similar accuracy.

Thanks to this prospective study, the hospital that conducted the study is now discussing how to commercially introduce this product.

These prospective studies will naturally spread to other regions within the country, and other European countries in the future.

Q: Are there any other studies that may receive attention during RSNA 2022?

A: We will also unveil a prospective study with Seoul National University Hospital regarding our chest software. This study compared the difference between when AI is used or not when patients use X-rays at health checkups in Korea.

Before the researchers conducted the research, the researchers involved in the study were worried about how much benefit AI would have.

This is because doctors look at X-ray results very sensitively when people do a health checkup in Korea.

Therefore, they were skeptical about how much more information an AI could find.

However, in conclusion, more meaningful pulmonary nodules could be found when using our AI system.

I think it is meaningful because it is a prospective study and a randomized controlled trial that compares the difference between using AI and not.

Q: I would like to ask about the business perspective of Lunit. During RSNA 2021, the company told KBR that it plans to focus more on the U.S. market from 2022. Has the company had any achievements so far?

A: In the case of the U.S., we have received U.S. Food and Drug Administration approval for our AI software for X-ray and 2D mammography.

In the case of 2D mammography, however, the market is very small, so it's a bit difficult for us to create something new, and now it's not a big deal in terms of investment.

Now that we have a product that covers 3D mammograms, we are again actively making investments, and we believe that this will help boost our presence in the U.S.

In the case of chest radiography, we are continuing our efforts with other global partners, such as GE, to see how we can establish a market in the U.S.

While we have little to show, for now, we are tightening our partnership with our global partners and trying to deliver improved results.