As medical artificial intelligence (AI) technologies gained attention amid the COVID-19 pandemic, latecomers have entered the market and begun to stand out, signaling that competition in the medical AI solutions market has entered a “second round.”

According to industry sources on the 6th, the early medical AI solutions market had been formed around leading companies such as Lunit, Vuno, and JLK. Recently, however, new players have been releasing AI medical solutions across various fields, including cervical cancer and stroke, achieving tangible results such as overseas exports.

These companies plan to capture the market by differentiating themselves from existing players, for example by applying AI diagnostic technologies based on pathological tissue or utilizing unsupervised learning. According to the Korea Technology and Information Promotion Agency for SMEs, the global smart healthcare market is expected to grow at an annual rate of over 20%, from approximately USD 96 billion in 2016 to about USD 206 billion in 2020.

doAI, a medical AI platform company, has developed a medical AI solution that diagnoses pathological tissue through microscopic interpretation, in contrast to most medical AI companies that focus on radiology-based AI diagnostic solutions. Even before the official product launch, the solution has drawn attention after being introduced in Cambodia.

With support from the Korea International Cooperation Agency (KOICA), doAI completed the development of an AI-based automatic interpretation system that analyzes cervical cancer using images obtained from microscopic examination of cell slides via a mobile app. The company has filed three related patents in Korea and the United States. It is also jointly developing a “chromosome alignment and interpretation AI solution” with the GC Green Cross Medical Foundation, which analyzes chromosome images during metaphase to determine numerical and structural chromosomal abnormalities.

In particular, the cervical cancer AI interpretation technology recently received approval from Cambodia’s Ministry of Health, after which remote video training sessions were conducted for Cambodian medical professionals.

A doAI official said, “Cambodia has an extremely small number of pathologists—only a single-digit figure—and culturally, women are not accustomed to undergoing medical examinations, so demand is high,” adding, “We plan to target developing countries with underdeveloped medical systems as the main market for our AI cervical cancer interpretation solution and to actively expand sales in the ASEAN region.”

The company was co-founded by CEO Choi Yong-jun and Professor Shin-Ho Do, a well-known expert in AI medical devices at Harvard Medical School. It has established partnerships with leading Korean hospitals, including Seoul National University Hospital, Asan Medical Center, Yonsei Severance Hospital, Seoul St. Mary’s Hospital, Yeouido St. Mary’s Hospital, Chung-Ang University Hospital, and Ewha Womans University Hospital.

AIDOT (formerly Buzzpole), an AI medical information and communications technology (ICT) company, developed the AI-based remote cervical cancer interpretation system “Cerviray AI” and exported it overseas. The company also developed a portable AI-powered carotid ultrasound device capable of pre-screening stroke risk, which received official approval from the “Clinical Research Information Service (CRIS)” for multi-institutional clinical testing. In recognition of its technological capabilities, the company was selected as one of the “Seoul Top 300 Promising Companies” in 2020, a program promoted by the Seoul Metropolitan Government and the Invest Seoul Center.

In March, the company signed a contract to exclusively supply its cervical cancer interpretation solution to Genius, a Swedish colposcope camera manufacturer. More recently, after signing a business cooperation agreement with the Chinese subsidiary of S-Net Systems to conduct sales activities in China, it also concluded an initial system export contract worth approximately USD 100,000, centered around Beijing.

AIDOT also expects to achieve export results of KRW 4 billion (USD 4 million) this year by supplying “Cerviray AI” to Southeast Asian countries.

Promedius, a medical AI solution developer spun off from the Medical Imaging Intelligence Realization Research Lab at Asan Medical Center in Seoul, plans to unveil a prototype of its solution for diagnosing abnormal regions in brain CT scans.

The company explained, “Most existing medical AI solutions use supervised learning, which trains on data and correct answers simultaneously, making it difficult to respond to rare diseases for which data are hard to obtain. In contrast, we utilize next-generation unsupervised learning-based AI, allowing us to overcome the limitations of existing AI models.”

Promedius was recently selected for the “2020 Early-Stage Startup Package,” which provides up to KRW 100 million in support from the Ministry of SMEs and Startups, and was also chosen for the 7th cohort of the “Startup Nest” program, with CNT Tech participating as an accelerator.

The solution under development is expected to be useful for tasks such as triaging emergency patients in hospitals, as it can pre-identify abnormalities across various brain lesions. Promedius also aims to supply customized medical AI software applicable to a wide range of clinical settings, including chest X-ray imaging.

Metro Newspaper / Reporter Chae Yoon-jung (echo@metroseoul.co.kr)