Yoichi Sato, Representative Director of iSurgery Co., Ltd.
Osteoporosis is a disease that makes bones fragile and easily break. Despite its high profile, it is said that less than 20% of osteoporosis patients over the age of 60 receive appropriate treatment interventions. We interviewed Dr. Yoichi Sato, an orthopedic surgeon and the representative of iSurgery Co., Ltd., which is promoting a project with AI in order to change this situation.
―What project is iSurgery Co., Ltd. working on?
We consider the lack of screening and treatment for osteoporosis a societal issue and are working to improve the rate of intervention. Specifically, our project aims for early detection of osteoporosis by analyzing data of plain chest radiographs with artificial intelligence (AI) and generating bone density reports. Plain chest radiographs are frequently taken in clinical settings and are used for lung cancer screening and regular medical examinations. We secondarily use them.
―Is it possible to judge osteoporosis only by plain chest radiographs without using specialized equipment?
Osteoporosis patients have their bone density measured in several parts of the body, such as the lower back, arms, and legs, and also have their plain chest radiographs taken when they visit a hospital or while being hospitalized. By training AI on both the data, we are developing an AI algorithm to predict bone density of the spine and femur from the chest radiographs. We can deduce normal bone density, low bone density, and suspected osteoporosis to an accuracy of 90%.
―Why did you decide to use plain chest radiographs?
Because they are frequently taken during general checkups, health screenings, and other medical services. We had our eye on the fact that a high percentage of people over the age of 40 had lung cancer screening and that there was high affinity with the age-specific prevalence of osteoporosis. In the first place, the rate of screening for osteoporosis is very low at about 5%, while that for lung cancer is about 50%.
―What process do you take to detect osteoporosis using plain chest radiographs taken at lung cancer screening?
In the case of a city with a population of 100,000, the number of lung cancer screening examinees over the age of 40 is estimated to be about 27,000. Therefore, about 27,000 people can be screened for osteoporosis by AI. In this case, about 3,300 new osteoporosis patients should be detected based on the age-specific prevalence of osteoporosis among people over the age of 40. Appropriate intervention in this population can prevent 107 people from fracturing their femurs, thereby contributing to a reduction in medical and nursing care expenses by about 200 million yen per year.
―Why did you decide to develop this kind of product and what were your concerns?
I have worked as an orthopedic surgeon for eight years since I became a doctor and have seen patients with osteoporosis. So many of them broke their bones because they had not received interventions.
I found that about 80% of people over 60 with osteoporosis did not receive interventions. Currently, the screening rate for osteoporosis is only 5%, and the intervention rate is about 10 to 20%. It is said that there are 13 million osteoporosis patients in Japan, but I suppose that the actual number is even greater if we include potential patients. We believe that early detection and treatment of osteoporosis is essential both for the medical economy of an aging society and for health promotion projects for the whole country.
―Why are we in this situation with intervention for osteoporosis?
The underlying issue is the low rate of bone density testing. This low rate is attributable both to medical providers and to patients.
Since devices to measure bone density is extremely expensive, the rate of DEXA (*2) installation is only about 6%. There is also a problem of space and personnel for installation. Meanwhile, examinees take the risk of osteoporosis less seriously than that of other diseases such as cancer, so it is difficult to motivate them to undergo osteoporosis screening. These made me think that we should integrate osteoporosis screening into an existing workflow.
―What is the current team structure of iSurgery Co., Ltd?
I started my company in May 2020, which now consists of five members: two doctors, one business strategist, one AI engineer, and myself. We are conducting R&D not simply for research purposes but with social implementation in mind. We team up with orthopedic surgeons from other institutions, data scientists from universities, and external supporters.
I have always liked challenges and problem-solving, and I considered that organizational action was necessary to resolve the big issue of increasing the intervention rate for osteoporosis. I feel that even if I cannot do anything by myself, I can do big things with a team of experts.
―How do you plan to proceed with the project in the future?
In terms of our product, we are developing AI by extracting data from a total of 20,000 radiographs from seven institutions. This year and next, we plan to conduct demonstration experiments with local governments and screening organizations, obtain pharmaceutical approval for a controlled medical device classified into Class II, and raise approximately 100 million yen from venture capital (VC) investors.
―What is your vision for the project?
It is said that there is a close association between osteoporosis and decreased ADL/QOL (3) as well as need for nursing care and the state of disuse (4). We will continue our activities with the aim of extending healthy life expectancy in an aging society through reliable intervention and continued treatment.
*1. The term “PMDA” is an abbreviation for Pharmaceuticals and Medical Devices Agency, an organization administered by the Ministry of Health, Labour and Welfare. It plays three roles: conducting approval review of pharmaceuticals and medical devices, ensuring safety measures, and providing relief services for adverse health effects.
*2. The term “DEXA” is an abbreviation for dual-energy X-ray absorptiometry, an X-ray bone densitometry device. It is globally designated as a standard measuring instrument in osteoporosis guidelines.
*3. The terms “ADL” and “QOL” are abbreviations for Activities of Daily Living and Quality of Life, respectively.
*4. The term “disuse” refers to a state in which excessive rest, such as being bedridden, causes a decrease in activity, resulting in various malfunctions, including mental and physical deterioration.
|Company Name: iSurgery Co., Ltd.|
|Number of employees: 5|
|Main businesses: Application development business, medical AI development business, and data science business|
This article is part of a series of articles introducing venture companies working together as ICF members to resolve societal issues.