China's healthcare insurance data has formed the world's largest and most comprehensive single-payer health database. The first national-level "AI + Imaging Recognition" professional competition will be held in Guangxi from August to October this year. For the first time, the National Healthcare Security Administration is globally soliciting AI solutions for medical imaging through the "National Healthcare Imaging AI Recognition Competition." Local governments will also provide supporting industry funds worth billions of yuan and a standardized dataset valued at tens of millions.
The medical AI imaging industry has long faced challenges such as "hospital-by-hospital deployment and lack of reimbursement pathways." Experts and industry insiders who attended the competition's promotional event in Beijing expressed hope that the administration's initiative would accelerate the interoperability of the healthcare imaging cloud platform and promote the inclusion of AI imaging in medical insurance pricing, thereby breaking down the final barriers to commercialization.
The competition is jointly organized by the National Healthcare Security Administration and the Guangxi Zhuang Autonomous Region Government. It focuses on AI technology applications and medical imaging-assisted diagnosis, inviting top-tier AI solutions from around the world. Guangxi will expedite the construction of its healthcare imaging cloud platform, aiming to build a standardized imaging dataset of 30 million cases and create at least five annotated datasets that comply with national medical data standards. Additionally, a 10-billion-yuan industrial guidance fund will be established to provide financial support for outstanding award-winning projects.
Cao Wenbo, Deputy Director of the Big Data Center under the National Healthcare Security Administration, stated that in recent years, the administration has been promoting the nationwide development of the healthcare imaging cloud. This year, it also launched a pilot program for the "Personal Healthcare Cloud," integrating medical and health data of insured individuals scattered across medical institutions, pharmacies, testing centers, and wearable devices.
As of March 25, the administration has collected 366 million medical imaging index entries. Currently, China's healthcare data covers over 1.33 billion insured individuals, forming the world's largest and most comprehensive single-payer health database.
The competition will closely interact with the "Personal Healthcare Cloud" initiative. Its eight tracks cover diseases with high incidence rates and significant disease burdens, such as lung cancer, kidney cancer, intracranial aneurysms, gliomas, prostate lesions, breast cancer, and thyroid cancer, as well as cutting-edge areas like multi-disease detection in chest X-rays and lesion recognition in ultrasound videos. Breakthroughs in AI-assisted diagnostic technologies for these tracks will directly contribute to the imaging recognition capabilities of the healthcare imaging cloud and the "Personal Healthcare Cloud" intelligent system.
Teng Gaojun, an academician of the Chinese Academy of Sciences and Director of the Medical and Life Sciences Department at Southeast University, emphasized that AI's current role in medical imaging is to assist doctors in preliminary screening rather than replacing them. AI can quickly locate lesions, significantly reducing the heavy workload of image interpretation.
"AI can help doctors quickly identify issues in their demanding work of reviewing hundreds of images daily, allowing them to focus on final diagnoses and shift to higher-value tasks," Teng said.
Chen Min, Director of the Medical Imaging Center at Beijing Hospital, noted that clinicians hope AI can shorten examination times while maintaining quality, and improve diagnostic efficiency and accuracy. AI can alleviate the burden of reviewing hundreds of images daily, reducing missed diagnoses caused by fatigue.
Despite over a decade of development, AI medical imaging still falls short of ideal expectations. Lu Xiaoliang, Vice President of iFlytek Co., Ltd., attributed this to two factors: rapidly evolving technologies like large AI models, and the previous reliance on fragmented, single-disease data, which is insufficient for developing multi-modal, multi-disease systems.
Current challenges in China's medical imaging industry include inconsistent image acquisition, quality, and diagnostic standards, as well as frequent redundant examinations.
Teng highlighted that the national healthcare imaging cloud platform promoted by the administration aims to enable free sharing and secure protection of imaging data, ensuring full control and real-time monitoring while addressing the lack of high-quality data in AI medical applications. The platform can set minimum imaging quality standards, provide timely feedback on subpar examinations, and lay a solid foundation for AI model training and clinical implementation.
The healthcare imaging cloud is transforming AI imaging deployment models. "Previously, AI imaging deployment required hospital-by-hospital integration. Now, with the cloud platform, AI capabilities can be deployed centrally, allowing doctors to use AI as easily as turning on a tap," Lu explained. The platform already interfaces with doctors' workstations across medical institutions, enabling AI companies to reach numerous hospitals by integrating with the cloud, significantly streamlining product deployment.
Zheng Chao, CTO of ShuKun Technology, emphasized that commercial companies are most concerned about whether AI can be included as a reimbursable item under medical insurance. Currently, some AI projects are listed as "extended items" or "additional charges." The future goal is for insurance to transition from "fee-for-service" to "value-based payment."
"Once the AI reimbursement system is established, companies can better serve all medical institutions and the public. We hope the competition's pilot projects in Guangxi will set an example for innovation and exploration in new technologies," Zheng said.
Guangxi has introduced several major support policies for the competition, including dedicated funding for award-winning projects, priority review and approval for Class II medical devices, up to 20 million yuan for key technology industrialization projects, a 10-billion-yuan AI industry fund, and policies such as open scenarios and "computing vouchers" to build an industrial ecosystem from research to application.
Next, the competition's promotional events will be held in Shanghai, Shenzhen, Chengdu, and select ASEAN countries, inviting domestic and international innovation teams to jointly advance AI technological innovation in medical imaging.