Artificial intelligence has worked its way into numerous business sectors under the guise of improving efficiency, but in the medical field it promises to do even more: save lives.
AI Medical Service, a Tokyo-based startup developing an AI-based endoscopic diagnostic support system, is convinced of the technology’s capabilities and aims to prove its effectiveness.
Gastrointestinal cancers are a major cause of cancer death worldwide — they account for 1 in 3 such deaths globally — and the startup says this is largely due to an inability to detect them at an early stage.
Data compiled last year by the National Cancer Center Japan show that the five-year survival rates for patients with Stage 1 stomach cancer was over 90%. But the rates plunge for those with Stage 3 or 4, standing at around 40% and 6.5%, respectively.
“Detecting the disease at an early stage is directly related to saving the patient's life,” said AI Medical Service CEO Tomohiro Tada, an endoscopist who founded the startup.
Cancer screening using endoscopes has been based on visual observation by humans, meaning subtle abnormalities might be overlooked, with the doctor’s experience being an important factor.
“We hope to contribute to a reduction in such oversights and to make the quality of medical care more equal” through the use of AI, Tada said.
On Monday, AI Medical Service launched its first AI software product, which is designed to improve the early detection of stomach cancer.
During an endoscopy, the AI system provides real-time analysis of the imaging data captured by the camera.
The system, which has received regulatory approval from the health ministry, can be integrated with existing endoscopes made by Olympus and Fujifilm, both of which make up a dominant share of the endoscope market globally.
AI offers an extra pair of eyes with a recognition level as good as experienced endoscopists, helping doctors conduct more thorough exams, AI Medical Service said. That will then help reduce the number of overlooked cases.
“Our AI system has been trained with an amount of data that a human doctor would not be able to learn in a lifetime. ... Working with such an AI system will make it possible to provide better medical care,” Tada said.
The startup said the AI system is more of an “assistant” for doctors, meaning it’s not the AI that makes the final decision on diagnosis, adding that this would nonetheless help alleviate the mental and physical burden on doctors.
The startup will keep upgrading the AI to improve its diagnostic accuracy and will continue its research and development to make the system capable of detecting other gastrointestinal cancers.
To provide diagnostic support to prevent other types of gastrointestinal cancer, the firm has also agreed to team up with Lpixel, a startup that provides image analysis technology for colonoscopies, to help detect colon polyps that might develop into cancer. AI Medical Service will market that system on behalf of Lpixel.
Tada, who founded his startup in 2017, was inspired to develop the AI diagnosis system after attending a lecture by Yutaka Matsuo, a professor at the University of Tokyo and renowned AI expert in Japan. There, Tada learned that AI’s image recognition skills were becoming superior to that of humans.
Tada believes that Japan can lead the field of AI-based endoscopy because the country is a pioneer of the medical procedure.
Olympus dominates the gastrointestinal endoscope market globally with about a 70% share, while Fujifilm is well-known as an endoscope-maker.
According to data compiled by Olympus last year, the number of endoscopists per 100,000 people in Japan stood at 28.2, considerably higher than other countries such as China, the Unites States and India, which had 2.8, 4.7 and 0.7, respectively.
“The quality and quantity of data is the most crucial thing for AI,” said Tada, adding that AI Medical Service has partnered with more than 100 medical institutions in Japan to secure endoscopy imaging data.
AI Medical Service has been selected for the economy ministry’s J-Startup program, which is designed to support startups that can be globally competitive.
The startup has already established bases in the U.S. and Singapore. Last month, AI Medical Service announced that the regulatory review and registration for its AI diagnostic system in Singapore has been completed.
“We will expand our business not only in Japan but also in Southeast Asia using our Singapore base. We are also looking to enter Brazil, the U.S. and other parts of the world,” Tada said.
“We are certain that expanding our products globally will help patients around the world.”
With your current subscription plan you can comment on stories. However, before writing your first comment, please create a display name in the Profile section of your subscriber account page.