How Japanese Startups Are Using AI to Transform Healthcare

In the quiet streets of Hiroshima, a small clinic has drawn nationwide attention for adopting six different types of AI-powered medical devices to assist in disease detection — an unprecedented number for a private facility in Japan.

One such system can detect abnormalities in electrocardiograms that would be difficult for the human eye to catch. Another uses endoscopic images to analyze conditions of the stomach and intestines, while a third focuses on the colon. These AI diagnostic tools are designed to identify signs of illness faster and more accurately than humans, alerting physicians to possible issues before they worsen.

The clinic’s 66-year-old director said he decided to introduce AI after realizing his own eyesight and physical stamina were declining. "My eyes and ears aren’t what they used to be," he explained. "I started worrying that I might overlook something during examinations, so I brought AI in right away. It really helps extend a doctor’s working life."

Across Japan, AI-assisted medical devices are becoming increasingly common. In August, for example, a software system that warns surgeons of areas that should not be cut during uterine surgery received regulatory approval. When paired with surgical-assist robots, it allows even less experienced doctors to perform operations with greater safety.

Among the fastest-spreading AI devices is Iris’s throat imaging system known as "Nodoca," approved two years ago for influenza testing. Now installed in more than 2,000 clinics nationwide, it is covered by health insurance, keeping costs comparable to conventional tests. The procedure is quick and painless: a small camera at the tip of a wand captures images deep inside the patient’s throat, and within seconds, the results are sent via the internet. Both doctors and patients say the system has greatly reduced their workload and discomfort.

"It used to involve swabbing the mouth, which made people gag," said one patient. "Now the nurse just takes a quick look with the camera — no scraping, no pain." Doctors also appreciate that nurses can handle the imaging while they simply review the results, freeing up time and resources.

Iris, the company behind Nodoca, was founded in 2017 and employs about 100 people. The firm’s CEO, 39-year-old Sho Okuyama, is a graduate of the University of Tokyo’s Faculty of Medicine and a practicing emergency physician. Inside Iris headquarters, Okuyama and his team are constantly exploring new possibilities for their throat-based diagnostic AI. During one recent internal discussion, engineers considered whether they could use throat images to estimate vital signs such as body temperature and heart rate. "If we can extract vital data from throat videos," Okuyama said, "it could open up an entirely new approach to screening."

Okuyama’s team believes that in the near future, Nodoca may be able to screen not only for influenza but also for COVID-19 and hypertension using the same throat images. Chief Technology Officer Fukuda attributes much of the company’s success to recruiting top global AI talent. "At first, I tried to build the model myself, but it didn’t go well," he recalled. "Then Yoshihara joined us, and within three weeks he identified the core problem. In just three months, he solved issues I couldn’t fix in half a year."

Yoshihara is one of only about 300 engineers worldwide recognized by Google as a top-tier AI expert. He also teaches as a research fellow at the University of Tokyo. "At first, our AI could only detect influenza," Yoshihara said, "but as we kept improving it, we found it could infer gender and even detect hypertension from throat images. The deeper we went, the more fascinating it became."

Iris now has five world-class AI engineers of similar caliber, creating a hub of exceptional talent. In late October, Okuyama traveled north to Monbetsu City in Hokkaido, known for its drifting ice. There, he also serves as a part-time emergency physician at Monbetsu General Hospital, working one weekend per month while managing Iris full-time.

"Having him here is a huge help," said the hospital’s chief surgeon. "Our surgical team is small, so when Okuyama is here, it really eases the load." Okuyama mainly handles night-time emergency cases. Asked why he continues to practice medicine despite his busy corporate role, he said, "Medical work has strict regulations, and if you want to change the system, you need to understand the field firsthand. Standing in the clinical front line gives me confidence and credibility. It’s one of the most important things I can do."

After graduating from the University of Tokyo, Okuyama began his career as an emergency doctor at the Japanese Red Cross Medical Center. He later worked on remote islands such as Teruma Island in Okinawa, where he experienced the harsh reality of emergency medicine in areas with limited medical infrastructure. "There were days when several patients couldn’t be saved. I often thought, if only we had noticed earlier or run a different test, maybe the outcome would have been different," he said. "That’s when I realized change requires more than treating patients one by one — it takes new technology and a broader approach."

Driven by that belief, Okuyama founded Iris with the goal of saving more lives through advanced AI medical tools. "If we can help prevent disease, even a little, that’s what I want to do," he said.

In early November, Okuyama welcomed a group of younger alumni from his alma mater to Iris headquarters, where he gave a lecture titled "Living as an Entrepreneur." He demonstrated one of his AI devices — a compact camera used in roughly 1,800 to 1,900 medical facilities across Japan — and emphasized the importance of collaboration. "To build good AI, you need cooperation from many hospitals and patients willing to let us collect images," he explained. "We also need to partner with academic societies and hospitals. Building a strong network of colleagues is absolutely essential."

Under Okuyama’s leadership, Iris has earned global recognition. Its Nodoca device won first place at an international startup competition last year, marking a major milestone for Japan’s growing medical AI industry. After hearing his lecture, one student said, "It was inspiring to hear how differently adults think about challenges compared to us students. It gave me a new perspective." Okuyama smiled and said, "I’m glad. This is one small way I can give back to my school."

Source: テレ東BIZ