China is already a powerhouse in AI for radiology and medical imaging. Next they're going global.

May 29, 2025

This is a video transcript, for the report found here:

Bullets:

A severe shortage of radiologists compelled China to introduce Artificial Intelligence across its hospital systems.

A comprehensive national effort began just three years ago, with the aim of developing a new industry for diagnostic medical imaging, all with AI.

Today, more than three fourths of Chinese hospitals rely heavily on AI systems. This has slashed costs for medical care delivery, improved patient outcomes, and dramatically reduced burnout among Chinese medical staff.

A crucial factor in the growth of China's AI is the vast biodata that researchers and tech companies have available to develop Large-Language Models for their systems.

China also enjoys access to enormous volumes of real-life patient data from around the world, as so much of the world's pharmaceutical R&D is performed here, by global companies.

That is a key differentiator and comparative advantage, which strongly positions Chinese medical AI in global markets.

Report:

Good morning.

Chinese industry is adopting AI and robotics much faster than any other country. The fact that China’s economy is growing much faster than its population is one of the drivers of this. But there is another factor, which is that Chinese AI is creating jobs which did not exist before. That’s to say that in the US and Europe, AI is replacing workers. Tech companies are laying off thousands of coders and engineers, for example, because so much of their work can be automated, and code can be written by AI programs. But in China, AI is being used to build workers that were never there in the first place.

A strong example of this is in medicine, in radiology. This report was published in the Korean Journal of Radiology. China never had nearly enough radiologists, and the problem is only growing. Tertiary hospitals are secondary hospitals, not the more advanced and developed ones, which serve major cities, and it’s to those hospitals where most of China’s patients go: 13% of the hospitals in China are tertiary-level hospitals, but they serve 60% of the population.

The burden on doctors, and radiologists, is very heavy, and the demand for radiology services and staff is growing much faster than medical schools can train new ones—over 7 times faster. So here are how those numbers look: one Chinese radiologist per a population of 70,000. Every day, one Chinese radiology doc needs to read over 300 X-rays, 80-100 CT scans, and 60-80 MRI scans. And there’s not enough time to get that done—even at 1 minute each, that’s a total of 480 minutes, which is 8 hours. And most of these reports require several minutes, 10 to 20 minutes of review, and reports, and consultations with the patient and the other doctors.

That volume represents a paralyzing problem, and means that millions of patients aren’t getting the care they need. Only heavy investment in AI, and in building LLM’s that can be deployed across all of China, which can do the work of tens of thousands of radiologists, can solve this problem.

So beginning in 2022, there was a strong push to build an industry that incorporates AI in medical imaging, and to build the hardware for it. Today, the growth in China’s diagnostic AI industry is growing exponentially. By last year, 2024, the penetration rate for AI was forecast to be 30-40%.

That report from the KJR was from 2023, for the estimates. In November of 2024, we see what the numbers actually were for last year. A survey of Chinese radiologists said that the actual penetration was 74.5%. Nearly half had been using AI in their jobs for over a year. The researchers pointed out years ago that radiologists in China were getting burned out—there isn’t enough time in a week to handle the work demands of even a single day—and as radiologists used AI more regularly, their work loads decreased, along with burnout. Now that problem is going away, as these AI programs are doing the work of thousands of doctors.

It’s interesting to compare those AI penetration rates, for Chinese hospitals, to what is going on in American hospitals. On the surface, it seems that US hospitals are doing the same thing. This headline from the NYT, and this report from Klas Research suggests a similar trend is underway in American clinics and labs. “AI adoption for medical imaging is soaring.”

That is from December 2024. But the data suggest that American radiology docs aren’t really doing it. More than half of healthcare organizations use at least one AI imagining tool, and those tools are commonly used in neurology, breast and lung cancers, and writing up reports. But the reality is that the penetration and use rates are very low. Most health care organizations – we don’t call them hospitals anymore--are using AI in just a handful of cases. 75% of respondents say their companies have only one or two patient cases, and fewer than 10% have five or more patients they have served with AI tools.

This pie chart suggests that there will not be much growth in the American market, either. 36% of users say they are “live” and considering more, 16% said they are live but no plans to do much more of it.

So the urgency to introduce new AI solutions into the American medical system is just not there, or at least nowhere to the extent that exists here in Asia. It’s possible that radiologists in American hospital systems are content with the industry as-is. Radiologists in the United States earn about 5 times what comparable radiology doctors make, here in China. And in China, remember, the workloads are far heavier.

But this poses another problem. China’s adoption of AI in radiology was borne of necessity, and national and provincial governments—along with Chinese universities and hospitals—adopted AI to create doctors where none existed. They did so, quickly, and Chinese AI systems are now well-positioned to dominate the market, globally, for this technology. Artificial Intelligence in diagnostic imaging and radiology, cancer detection and treatment, telemedicine—these are advanced applications that China will be exporting to the rest of the world. These are trillion-dollar industries, and China is already far ahead.

A big part of the reason why, is right here: China has strict policies that restrict the sharing of Chinese patient information, to foreign companies. But China does have access to biodata from patient populations elsewhere. The datasets from Chinese patients are enormous, and Chinese researchers can use those, freely, along with the data from patients across the world. Remember too that European and American pharmaceutical companies outsource most of their research and development to Chinese companies, and it’s here where most of the Active Pharmaceutical Ingredients are made.

This report goes deeper still, China has access to biodata from around the world, but its medical data are closed off. More biodata means that Chinese companies can build more powerful medical AI.

Now Beijing wants to commercialize medical AI, and Chinese tech companies that were building AI for industrial applications are getting into medicine. China has a centralized strategy, and biodata is a resource to be used to advance their AI industry. China next will bring these technologies to customers outside China, and the fact that Chinese companies have so much more access to so many more live patient medical data is a big comparative advantage. “China could become a global powerhouse in medical AI and capture the economic value” that goes along with it.

Resources and Links:

America Is Missing The New Labor Economy – Robotics Part 1

https://semianalysis.com/2025/03/11/america-is-missing-the-new-labor-economy-robotics-part-1

Bloomberg, Microsoft Layoffs Hit Coders Hardest With AI Costs on the Rise

https://www.bloomberg.com/news/articles/2025-05-14/microsoft-layoffs-hit-software-engineers-as-industry-touts-ai-savings

The AI layoffs begin

https://www.reddit.com/r/ChatGPT/comments/1kqdukp/the_ai_layoffs_begin/#lightbox

Salary.com, Diagnostic Radiologist

https://www.salary.com/research/salary/alternate/radiologist-diagnostic-salary

Salary Expert, Radiologist in China, Mean Salary

https://www.salaryexpert.com/salary/job/radiologist/china

AI adoption for medical imaging soars: Klas Research