VUNO: Pioneers in medical AI solutions

In advanced economies like South Korea, Japan and the United States, aging populations are not only creating a significant social burden but also contributing to a shrinking labor force. Technology and medical advancements have been proposed as potential solutions to both extend life expectancy and improve life quality. How much, in your view, can technological and medical innovations help alleviate the impact of an aging population on the global economy?

You raise a wide-ranging question that pertains to a global scenario. Indeed, like Japan, South Korea is progressing into an aging society, with projections showing that around 10 to 20 percent of its population will be over 65 years old in the upcoming years. This demographic shift significantly impacts the labor force.

From our perspective, a major issue surfaces in the form of a demand and supply imbalance within the medical industry. As societies age, there will be an escalating demand for healthcare services. However, training new doctors is a time-consuming process, creating a disparity between supply and demand.

Our objective is to employ technology to bridge this gap efficiently, primarily through the integration of medical AI. Various digital enterprises are harnessing technology to minimize this demand and supply discrepancy. For instance, with the aid of accurate AI analysis and diagnosis, a single doctor could potentially cater to more patients without compromising the accuracy of their diagnosis. The goal is to maintain the same quality of service while increasing efficiency.

At the outset, AI technology was primarily used for analysis and assessment, such as scanning CT, X-ray, or MRI results. This application helped doctors quickly analyze data, enabling them to detect diseases promptly.

The subsequent phase saw the use of AI extend to MDs or managers, helping them understand how to interpret imaging results, which proved incredibly beneficial. Following this, AI technology began to address the mismatch between supply and demand, with a focus on the treatment sector.

For instance, AI can decide the appropriate medication for treating cancer patients. However, more advanced AI is necessary to tackle the challenges we're currently facing, such as older populations living in dispersed, rural areas. Rural areas often lack adequate medical facilities, and accessibility is usually a problem. Therefore, it's crucial to have a system that allows for early disease detection, such as myocardial infarction, to prevent further complications. This is essentially aiming for faster, safer, and self-regulated health management. This is the direction we're heading in as an AI company, and it is the field we intend to expand our business into.

In recent years, the Korean Government unlocked new funding and resources to assist the development and expansion of the domestic medical equipment sector. How has this increase in government support affected VUNO?

Since our establishment in 2014, it took us four years to introduce VUNO Med-BoneAge in 2018. At that time, registering these medical devices with the government required going through approvals and regulations. However, there was no specific guideline for these software-based medical devices before 2017.

Things started changing when the new government shifted its focus towards the bio and medical industry, moving away from semiconductors and manufacturing. This resulted in a relaxation of many regulations and increased government support for startups like ours.

For instance, last year, around 130 AI-related devices were launched into the market. Once these devices are commercialized, the challenge lies in successfully selling them to the market.

As you mentioned about ecosystems, obtaining clinical data is crucial. More and more hospitals are using our devices to leverage their data, thereby creating a positive feedback loop and fostering a constructive ecosystem. In the past, the focus was more on hardware devices such as CT or MRI machines, most of which were imported from famous overseas brands like Philips. At that time, we didn't have much regulatory information or knowledge about device registration, since most of the clinical data was coming from those companies.

However, the supportive stance of the current government has created new opportunities for software medical device manufacturers like us. We can introduce our products to the market more easily now.

Nevertheless, every time we launch a new product, we must undergo clinical testing to ensure it can be effectively utilized by hospitals. We also have to determine the right price range for hospitals to use our devices. Our aim is to build a more stable software ecosystem that can support doctors and hospitals in effectively utilizing our new technologies.

As Korean brands begin to dominate the domestic market and increase their international presence in medical technology, how would you evaluate the current competitiveness of the Korean ecosystem, including startups, private sector, and government? How capable is this ecosystem in competing globally, particularly in mature markets like the U.S. or Europe, considering the importance of software-hardware integration in achieving global competitiveness?

To be candid, the competition in the field of hardware medical devices is formidable. Foreign companies have amassed a significant amount of clinical data, a key factor for reliability since medical devices are intrinsically linked to life-saving measures. While Korea is expanding its market share in areas such as ultrasonic waves and X-rays, competing with established foreign brands remains challenging due to their extensive database from prior performance.

However, when it comes to software, we perceive substantial opportunities. Although high-quality image and data capturing devices already exist in foreign markets, the focus is increasingly shifting towards software-based solutions. These solutions are key for determining the accuracy with which doctors can interpret the data. It's worth noting that human interpretation is not infallible, and doctors' judgement can vary based on their competence or even their state of well-being. For instance, after a long night, a doctor might struggle to read and analyze high-quality images or videos. This issue became even more pronounced during the Covid-19 pandemic, when doctors had to shoulder even more responsibilities due to labor shortages. Therefore, the accuracy of data interpretation is now of greater importance than the quality of objective data. With this shift in the medical landscape, AI technology is emerging as a game changer, creating more business opportunities for companies like ours.

Focusing on the Korean scenario, I believe the environment is highly favorable for the development of software-based medical devices. We have a host of skilled doctors, accessible top-tier hospitals, a high-quality medical system, and top-notch AI technologies facilitated by Artificial Intelligence clusters. Moreover, there's PACS, where we accumulate clinical data. Consequently, we have an immense amount of high-quality clinical data that can be meticulously and accurately analyzed by advanced AI technology. Given these factors, I am confident that the Korean software-based medical industry is highly competitive in the global market.

We also have access to a significant amount of clinical data from overseas. Korean doctors and researchers are actively publishing medical papers in international journals. Furthermore, Korean companies' participation in the Radiological Society of North America (RSNA) has been on the rise, demonstrating the growing recognition of Korean software-based medical companies. To sum up, while the Korean software-based medical industry is still in its early stages, we are well-positioned to enhance our competitiveness and increase our market share.

As the use of AI in the medical field develops, significant concerns around ethics and accountability, especially in cases of potential misdiagnosis by AI, are emerging. Understanding the AI's role in valuing human life and deciding treatments, as well as determining who is accountable when AI errs, indicates a need for robust legal and regulatory frameworks. Could you share your thoughts on these challenges and propose potential solutions to facilitate broader adoption of AI technologies in healthcare?

This is indeed a complex issue and what I share is more of a personal viewpoint rather than a definitive solution. I believe that efficacy should be prioritized over ethicality in this context. In medicine, there is a concept called "appropriateness." For instance, if there's a yet-to-be-approved drug that has a 50% chance of saving a life, even amidst uncertainty, I would opt for its use, given the alternative is certain death. The reality we must confront is the pronounced mismatch between the number of competent doctors available and the patient load they are burdened with, a situation exacerbated by the COVID-19 pandemic. Doctors can only handle a limited number of patients before they experience burnout. This mismatch is not exclusive to developing countries but also prominent in advanced nations such as the US and Europe, particularly in remote areas with inadequate medical infrastructure.

Indeed, high-quality medical devices can provide accurate images, data, and videos, but without a doctor present to interpret these, the devices serve little purpose. This is where AI enters the picture, providing initial analysis and assessment of data, and advising patients when they need to consult a doctor. While it may not be perfect, it does improve the odds. The goal of integrating AI into healthcare is to reduce medical disparities, improve survival rates, and enhance the quality of medical services.

Consider this scenario: in every 1,000 hospitalized patients, two to three might experience cardiac arrest. Owing to severe staff shortages in general hospitals, each nurse is responsible for a multitude of patients, and doctors can only make rounds approximately every eight hours, unless they are in the ICU. Consequently, patients in general wards are more likely to die from cardiac arrest as they remain unattended during the crucial period immediately following the cardiac event. In such instances, our AI device is invaluable, as it can predict the risk of cardiac arrest within 24 hours by monitoring vital signs such as heart rate, blood pressure, temperature and respiration rate. This facilitates early detection and improves patient outcomes.

Given that doctors cannot simultaneously attend to each patient, AI can at the very least assist in this respect. The use of AI is now extending beyond diagnostics to screening, early detection, and health management, blurring the boundaries between sectors. Our challenge lies in collaborating with policymakers, civil society, and the government to determine how AI can be effectively integrated into healthcare, without overlooking the pivotal issue of data responsibility. As AI is currently employed to predict future patient outcomes, it's crucial to garner doctors' trust.

VUNO was founded in 2014. At the time, the use of AI in the medical field was at an embryonic stage, so what motivated you to start this venture?

I began my career with a background in computer engineering, initially working at the Samsung Advanced Institute of Technology, an organization renowned for its forward-thinking approach. This experience granted me an early insight into the future potential of technologies like deep learning and AI.

During my time at Samsung, my primary focus was on voice recognition solutions utilizing deep learning and AI. Even in noisy environments, we were able to achieve impressive results, accurately recognizing voices and sounds. This experience showed me the potential of these technologies. It's interesting to note that voice recognition technology itself has roots dating back to the 1950s, and while there have been steady improvements over the years, there was a notable exponential advancement at a certain point.

Additionally, while at Samsung, I encountered the Picture Archiving and Communication System (PACS) used in hospitals, which converted Excel files into data formats. High-quality data input, including clinical data, was vital for high-quality outputs.

These experiences and insights made me realize the potential of blending AI technology with the medical sector, and how it could significantly improve people's health and safety. This is what inspired me to venture into creating a company that leverages AI for healthcare in 2014, despite these technologies not yet being mainstream.

VUNO achieved incredible sales growth, rising from 82 million KRW to 8.2 billion KRW in four years. The last quarter(2022 4Q) alone saw sales reach 6.3 billion KRW. Could you share insights into the key factors driving this growth and your strategies for consistently increasing revenues over time?

I can certainly provide some insight into our business growth. We entered the KOSDAQ in February 2021, recognizing that the expanding medical industry presented a significant market opportunity. However, the outbreak of Covid-19 presented a roadblock to our overseas expansion plans.

Thankfully, the government has been highly supportive of companies like ours, especially those engaged in AI-related healthcare and medical industries. We've received much positive feedback for our AI solutions, which has certainly contributed to our growth.

Additionally, we're proud to be the first software-based medical company to not only receive official certification but also gain the rights for billing patients who use our AI-based cardiac arrest prediction device, VUNO Med-DeepCARS. This has undeniably fueled our sales growth.

Furthermore, we're currently pursuing certifications like FDA, CE, and PMDA to expand our markets to the US, EU, and Japan. While FDA process is ongoing, we have accomplished a significant amount. By the end of this year, we expect to see growth in overseas sales, further boosting our CAGR.

Your product, VUNO Med, aims to enhance the interpretation and analysis of data from various electronic monitoring devices. Could you discuss the impact of these products on diagnostic accuracy by physicians? Have you noted any noticeable enhancements in their productivity when using your products?

We have gathered clinical data that suggests patients' health outcomes improved by double-digits when software-based medical devices were utilized, compared to when they weren't. For the VUNO Med-BoneAge solution specifically, while experienced medical professionals may not require it, it's been a significant aid for general practitioners. Even more dramatically, the benefits are amplified for non-specialists.

Beyond the numbers, a significant advantage of software-based solutions is the speed and consistency with which data can be analyzed. This is independent of the doctor's personal circumstances, which can often be subject to error and variability. Usually, when diagnosing a patient, the process involves progressing from an X-ray to a CT and then an MRI scan. Our aim is to provide a comprehensive lineup of these devices to aid doctors in detecting nodules and determining whether they're malignant or benign. By using both X-ray and CT, we're streamlining doctors' workflow, aiding them in identifying nodules from images.

Another noteworthy product we offer is VUNO Med-DeepCARS, an AI-based solution predicting cardiac arrest within 24 hours. This tool assists nurses and doctors by providing them with crucial data, enabling more thorough patient care. Hospitals using VUNO Med-DeepCARS have given positive feedback, as the solution addresses a fundamental problem: it not only allows medical professionals to manage all patients effectively but also helps identify those at higher risk of developing severe conditions. The key advantage here is that it lightens the workload of the healthcare staff while enhancing their work efficiency.

Could you elaborate on your global expansion strategy for VUNO Med, particularly in regions like Europe and the United States?

We are collaborating with the CHC group in Taiwan to export chest X-ray technologies, and we plan to venture into CT and MRI sectors as well. As for expanding into overseas markets, the initial step involves obtaining requisite approvals from health authorities like the FDA (U.S), CE (Europe), or ANVISA (Brazil). Unlike established products, we are in a situation where we need to gain approvals that include clinical trials, and VUNO Med-Lung CT AI is one such product we're working on for U.S. and EU approval.

However, gaining these approvals isn't the only critical aspect. Demonstrating the effectiveness of our devices is equally important. For instance, we must convincingly demonstrate that our VUNO Med-DeepCARS technology has indeed been successful in predicting patients' risk of cardiac arrest. It's crucial to build a substantial statistical case that our devices have effectively contributed to reducing the number of cardiac arrest patients. In this regard, we are intensifying our efforts to compile meaningful clinical data by teaming up with significant U.S. hospitals while also seeking approvals in different countries.

So far, we have won CE approval for five devices and are currently awaiting FDA approval. Moreover, our lung CT device has received approval from Japan's PMDA.

VUNO Care is a platform that incorporates three different hardware products, ranging from ECG to blood pressure and temperature monitoring, all consolidated into a single platform. Could you shed some light on the concept underlying this new segment and the specific goals you aim to accomplish through VUNOcare?

Earlier in our conversation, I touched upon how AI technology initially focused on diagnostics but has since expanded to include treatment, prevention, and health management. Essentially, VUNO Care is a platform designed to empower patients, enhancing their self-care and health management skills before an illness progresses into a serious condition. It's quite similar to how patients with chronic conditions like high blood pressure or diabetes manage their health in everyday life. Blood pressure can be monitored with a gauge, and diabetes can be assessed with a glucose monitoring device like an FGM. Essentially, effective management of health conditions requires measurable data.

Similarly, heart diseases like heart failure are also chronic conditions requiring continuous attention, but until now, there hasn't been a comprehensive system for managing these conditions. This was largely due to a lack of data necessary for assessing, analyzing, and monitoring patients with heart diseases. We identified this gap and decided to focus on developing measurement tools for heart diseases by integrating AI technology. The result was the creation of our medical devices, which are personalized and tailored to meet the needs of our patients.