According to a research report, as of the end of June 2026, the national medical insurance imaging cloud index had reached nearly 440 million entries, and five provincial-level regions had completed procurement tenders for their provincial imaging clouds. The report suggests that the medical insurance imaging cloud will propel the industry's business model to evolve from single-hospital software projects towards provincial platform development, ongoing operation and maintenance, and value-added application services. The revenue model for AI imaging is also expected to shift from one-time project fees to software licensing, subscription models, and payment based on examination volume. As the marginal scarcity of product registrations diminishes, industry competition is increasingly focusing on implementation capabilities. Companies with strengths in equipment synergy, regional platform development, and cross-selling across multiple disease areas are poised to achieve sustained revenue growth with lower marginal customer acquisition costs. The main viewpoints are as follows:

Data Infrastructure Gradually Improves, Unlocking Application Space for AI Imaging

As of the end of June 2026, nearly 440 million medical insurance imaging cloud index entries had been uploaded cumulatively from 31 provinces and the Xinjiang Production and Construction Corps. Five provincial-level regions, including Guizhou, Xinjiang, and Guangxi, have completed procurement tenders for their provincial imaging clouds, while areas like Tianjin, Inner Mongolia, and Hainan are progressively advancing their initiatives. With the gradual standardization of imaging data storage, access, and interface protocols, the data sources for AI imaging are expected to expand from single-hospital data to regional imaging platforms. Application scenarios are also set to extend beyond single-disease auxiliary diagnosis, such as for pulmonary nodules or fractures, to encompass cross-hospital data access, imaging quality control, identification of duplicate examinations, and comprehensive analysis of multiple diseases. The report posits that the medical insurance imaging cloud will drive the industry's business model upgrade from single-hospital software projects to provincial platform construction, continuous operation and maintenance, and value-added services. The industry's value chain is anticipated to spread from mere algorithm licensing to encompass data governance, PACS system upgrades, cloud resources, and AI application platforms.

Hospital Application and Payment Loop Gradually Connect, AI Imaging Revenue Model Poised for Upgrade

For AI imaging to achieve commercial scale, three key links must be sequentially established: imaging data access, clinical adoption in hospitals, and medical service charging. On the application front, as of the end of March 2026, 85 designated medical institutions in Jinan had accessed the medical insurance imaging cloud, cumulatively collecting and uploading 8.07 million imaging examination records. This has formed a business loop encompassing imaging data collection, index upload, patient access, and direct medical insurance settlement, providing a foundation for the large-scale integration of AI algorithms into clinical workflows. On the payment side, radiology examination pricing policies have begun to include AI-assisted diagnostic technologies, such as rapid medical imaging diagnostics, as expandable price items. This indicates a shift for AI imaging from being a hospital informatization cost to a quantifiable medical service value. The report believes that as hospital access scales up and local charging policies are implemented, the AI imaging revenue model is expected to transition gradually from free trials, research collaborations, and one-time project construction towards software licensing, annual subscriptions, and payment based on examination volume. Key performance indicators will also shift from the number of trial hospitals to metrics like paid examination volume, price per service, renewal rates, and payment cycles.

Product Supply Rapidly Expands, Industry Competition Shifts Focus to Implementation Capability

Newly approved application scenarios for AI imaging have expanded from mature areas like CT and DR to nuclear medicine PET-CT, reproductive genetic pathology, digestive endoscopy, cervical pathology, and primary-level tuberculosis screening. In June, United Imaging Intelligence's chest nuclear medicine image-assisted triage software obtained the first Class III medical device registration certificate in China's nuclear medicine AI field, bringing the company's total number of NMPA Class III-certified AI applications to 20. Shukun Technology holds 19 such certificates, indicating that leading companies are progressively完善 their product portfolios. The report argues that as the marginal scarcity of registration certificates decreases, industry competition will depend more on factors like hospital access, adaptation to clinical workflows, establishment of payment loops, and large-scale delivery capabilities. Investment evaluation should correspondingly shift from the number of certificates obtained to metrics such as the number of paying hospitals, examination volume per hospital, and renewal rates. Companies with capabilities in equipment synergy, regional platform development, and cross-selling across multiple disease areas are likely to drive sustained revenue realization with lower marginal customer acquisition costs.

Risks include potential delays or underperformance in the procurement and implementation of medical insurance imaging clouds and local charging policies; slower-than-expected hospital adoption, clinical use, and payment conversion for AI products; and pressures on medical institutions' informatization investments and payment collections.