Lunit is a medical AI company devoted to developing advanced medical image analytics and novel imaging biomarkers with its cutting-edge deep learning technology. The company has been internationally acknowledged for its advanced, state-of-the-art technology and application in analyzing medical images.
Lunit specializes in developing medical AI software products that can foster accurate and cost-efficient diagnosis and treatment. By providing an accurate quantitative assessment of disease, products are designed to significantly reduce unnecessary tests, treatment, especially expensive invasive tests, such as biopsies. Currently, the company has released two products that provide AI analysis for chest X-rays and mammograms. The accuracy level of their diagnostic products currently ranges from 97% to 99%, depending on the area of diagnosis.
Lunit’s Commitment Since Inception
Lunit was founded in 2013 by six computer science/engineering experts from KAIST, one of the most prestigious academic institutions in Korea. The company’s initial trigger was to provide AI analysis in the area of fashion. As distant as it may seem, back then, AI technology had not been as developed as to how we see nowadays. Lunit team quickly foresaw that their AI technology in development had much bigger potential, adaptable into more accurate and challenging missions which go beyond the fashion industry.
In medicine, on the other hand, a mere increase in performance by 1% can save many lives. It is the area that requires the utmost accuracy and expertise, and thus Lunit figured, if AI technology is applied in medicine, it can be utilized with great value in comparison to other industry areas.
Since then, the company has been continuing their journey successfully into the world of medical AI, under the mission to empower physicians with clinically actionable insights. Through AI, Lunit rediscovers and defines important findings in medical images—in order to help clinicians, make cost-efficient diagnosis and treatment and calls this technology “Data-driven Imaging Biomarker.”
A Strong Leadership Team
Lunit’s Co-founder and Inaugural CEO, Anthony Paek, who is now company’s Executive Chairman, had largely contributed in setting the coordinates and the course of its journey. He had successfully led Lunit from the scratch: planting the seed, gathering investments, and leading growth into where they are now. Anthony’s passion for deep learning as an electrical engineer had determined Lunit’s early focus on AI and the followed decision to delve into “medical AI”; his vision to develop DIB (Data-driven Imaging Biomarkers) that goes beyond human vision is where the company’s eyes are still fixed on.
Lunit’s current CEO, Brandon Suh, has been working with Lunit since 2016 as Chief Medical Officer, spearheading their efforts to elevate medical expertise. This year in October, he was appointed as CEO, opening a new phase of their fast-growing business: highlighting their professional capacity and identity as a “medical” AI company and accelerating overseas partnerships and expansion. Brandon’s background as a medical doctor and his proven track record in both medical expertise and business management have been enormously helpful in elevating Lunit’s presence and influence in the global medical AI sector.
Awards and Recognitions
In 2017, Lunit was selected as “AI 100 Startups” by CB Insights and as “Top 5 AI startups for social impact” by NVIDIA; the company’s studies have been published in major international publications such as Radiology, Scientific Reports, Clinical Infectious Diseases, and etc., Lunit’s AI/deep learning technology has won top places in international computer vision competitions such as ImageNet 2015, MICCAI TUPAC 2016, and CAMELYON 2017, surpassing top companies like Google, IBM, and Microsoft.
The company has tested its two most mature products, Lunit INSIGHT for Chest Radiography and Lunit INSIGHT for Mammography in various settings in 2018. Lunit has an online demo site where anyone can upload images and check out the results its algorithm brings within seconds. Additionally, Lunit has established various reference sites throughout the world. Overall, in 2018, the company has tested its algorithms in more than 1.5 million cases from more than 70 countries. The feedback so far has been terrific, and many those who have tried multiple similar products have commented that their solutions perform the best, especially in terms of the level of accuracy.
Transforming Ideas into Cutting-Edge Solutions
The company believes the key feature is the depth of information. More information is available and that information is being interpreted more effectively, leading to an improved utility and innovation.
“Our AI-centric approach is what makes our technology exceptional: our research team is committed to addressing the fundamental challenges of AI, defining problems from domain-specific issues and solving it in our own, creative way. We are in constant searching for novel ideas and solutions, “pushing the boundary.” We believe, at the end of the day, our day-to-day efforts will draw the future trajectory of the industry,” said Anthony.
Collaborations Driving New Success
“Perfecting Intelligence, Transforming Medicine” is Lunit’s corporate vision. The company is constantly challenging the boundaries of AI to achieve super performance, going well beyond the level of human vision. Not only is Lunit focused on developing their core technology, but also puts a great deal of effort into collecting the right data. For example, for the company’s chest X-ray product, the team has collected and used the ground truth labels from paired CT images, not from radiology reports or annotations from radiologists. Although it was difficult to find large-scale data of chest X-rays paired with CT, this approach enabled the organization’s algorithms to perform well beyond the level of human experts.
Recently, Lunit team has announced their corporate partnership with FUJIFILM. The team has also been working with Philips through the Philips HealthWorks program. The company is also actively engaged with various major vendors in both the radiology and the digital pathology space. Additionally, according to Lunit, new exciting announcements are yet to come in the coming months and throughout next year.
Embracing Industry Challenges
“Regulatory approval is always a big challenge for companies in medicine/healthcare, especially for startups like Lunit. It is difficult because, while the process takes a long time and significant resources, there are no specific manuals or guidelines for this, so there’s a great deal of uncertainty involved in the process,” said Brandon on the challenges faced by healthcare organizations.
Making the products seamlessly integrated into the clinical workflow of their users, the doctors are another challenge Lunit has faced. Establishing a strategic partnership with major medical device vendors is the key, the company believes, and that’s always a challenging task.
Brandon believes that starting from workflow optimization and improvement in efficiency, artificial intelligence in medicine will significantly augment the capabilities of doctors by providing them with deeper insights into patients. It will become an integral part of precision medicine, where diagnosis and treatment can be tailored for each individual, leading to profoundly more effective diagnosis and treatment, with better clinical outcomes and reduced medical costs.