China scans more patients, takes more X-rays, and generates more medical images than any country on earth. That data — now exceeding a billion annotated medical images — has become the training substrate for an extraordinary generation of medical AI systems. Companies like Infervision, Airdoc, and Yitu Healthcare are building diagnostic tools that, in carefully controlled studies, match or exceed the performance of experienced radiologists in specific tasks.
What AI Can Already Do
In clinical settings across China, AI systems are now routinely screening chest CTs for lung nodules, analyzing fundus photographs for diabetic retinopathy, and flagging cervical cancer precursors in pathology slides. These are not research demonstrations — they are systems deployed at scale, processing millions of cases per year, providing a second-check layer that is catching cancers that human readers might have missed on an exhausted Friday afternoon.
The results in population health screening have been particularly striking. China's national lung cancer screening program, which uses AI to prioritize high-risk CT findings, has detected early-stage cancers at rates that significantly exceed what human screening alone achieved. In a disease where early detection can be the difference between a cure and a terminal diagnosis, this is not an abstract improvement — it is lives saved.
