According to reports, RIMAG GROUP (02522) saw a surge of over 6% in early trading, with the stock currently up 6.23% at HKD 17.23 and a trading volume of HKD 80.6296 million. The movement follows the company's announcement regarding its participation in the launch of AIR, a path-level AI-assisted diagnostic product for chest CT scans, developed by its invested startup, Shanghai Yinghe Yimai Intelligent Technology Co., Ltd. (Yinghe Yimai).
On October 16, 2025, Yinghe Yimai officially released AIR, marking a significant leap in medical imaging AI technology from a "single-disease small model" 1.0 era to the "foundation large model-driven" 2.0 era. This product boasts capabilities for comprehensive diagnosis across all lesions and diseases in multiple organs, representing a milestone breakthrough in intelligent medical imaging.
The AIR product operates based on scan areas for diagnosis and the first roll-out focuses on chest CT scans, allowing for the simultaneous detection and diagnosis of lesions in the lungs, pleura, and mediastinum with a single scan, covering all common diseases in that area. Developed using the world's first cross-modal, full-process medical imaging foundation large model "Yinghe Miyi®", AIR includes multi-organ interactivity capabilities, showcasing its efficiency in diagnosis, precise coverage, interactive coordination, and seamless integration.
More than just a diagnostic tool, the AIR product is the first commercialization outcome of Yinghe Yimai's foundation large model ecosystem. By connecting top-tier hospitals and primary healthcare institutions through open interfaces, it establishes a positive feedback loop of "data-fueled model iteration and model-driven scenario implementation." This approach is expected to significantly shorten AI product development cycles, reduce costs, and help healthcare organizations achieve standardized screenings, intelligent diagnostics, and accessible services, effectively promoting industry chain cooperation and constructing a closed-loop ecosystem of "data-model-scenario."