On August 5, Frost & Sullivan officially released the "White Paper on the Global Digital Intelligent Medical Imaging Ecosystem Industry in the AGI Era." This white paper systematically outlines the underlying logic of how AGI technology reconstructs medical imaging services, clinical diagnosis and treatment, and ecological collaboration, providing empirical evidence in typical scenarios such as cranial, chest, and knee examinations. It identifies Yinghe Medical Imaging as a representative technological force within China's digital intelligent medical imaging ecosystem and explicitly states that RIMAG GROUP and Yinghe Medical Imaging "collectively form the core closed-loop capability within the digital intelligent imaging ecosystem, making this full-stack closed-loop model unique globally."

For institutional investors in RIMAG GROUP (02522), this white paper provides a cognitive framework to reassess its valuation logic. The "scenario-data-standard-model-product-feedback" full-stack closed loop built by Yinghe Medical Imaging is opening a cognitive pathway for the parent company to transition from a service industry valuation to a tech platform valuation. Understanding the internal logic of this valuation leap requires dissecting three progressive core propositions: First, what is the redefinition of RIMAG GROUP's offline assets in the AGI era? Second, why does the technological path of Yinghe Medical Imaging constitute a rare and difficult-to-replicate barrier? Third, how will the valuation mapping of this combination drive the cognitive restructuring of the capital market?

Asset Upgrade: Strategic Shift of the Offline Imaging Network from "Service Capacity" to "Data Infrastructure"

The traditional valuation logic for RIMAG GROUP has long been anchored to its identity as "China's first listed company in third-party medical imaging services." As of the end of 2025, the company has established a network of 117 imaging centers across 20 provinces, autonomous regions, and municipalities. Previously, this heavy asset layout was viewed linearly, with a clear pricing logic for such models in the capital market, but equally clear limitations—the number of core prefecture-level cities in China that can support the independent operation of imaging centers is ultimately finite.

However, this cognitive framework is being completely rewritten by the AGI era. When the 117 imaging centers are no longer seen as isolated service terminals but as a distributed network continuously generating real clinical scenario data, their strategic significance transcends the old definition of "inspection service capacity." The white paper clearly positions RIMAG GROUP as the "scenario and data cornerstone of the ecosystem"—"owning China's largest and most diverse third-party imaging center network," "not only providing a solid offline inspection service scenario but also leading the establishment of the industry's leading, full-modal imaging data standard system covering CT, MR, DR, ultrasound, and nuclear medicine." This standard system acts as the "common language" of the ecosystem, ensuring that massive medical data can be deposited and managed with high quality and standardization, providing indispensable "fuel" for AI evolution.

The value of data is not just in "having it," but in "being usable." According to the last disclosed reporting period, leveraging a data foundation of over 28.53 million imaging cases, a daily increase of 20,000 cases, and full coverage across 12 modalities, RIMAG GROUP pioneered the construction of a complete closed loop from data resources to data assets to data transactions. The true competitive barrier is the data standardization project the company has been strategically deploying since 2021—the "Medical Imaging Examination Item Name and Coding Standard" released in 2024, covering 11 major coding systems, upgrading data from "raw accumulation" to "trainable AI fuel." This standard has become an important reference blueprint for the National Healthcare Security Administration's "Medical Insurance Imaging Cloud Index," and multiple data products have been listed on the Shanghai and Beijing data exchanges, generating tiered data service revenue within various local government data trusted spaces.

From a capital market perspective, this asset upgrade is reshaping the valuation base of RIMAG GROUP. When data transitions from a "cost center" to a "revenue center," and the 117 centers evolve from "service terminals" to "data infrastructure," the "global leader in the digital intelligent imaging + ecosystem" vision touted by RIMAG GROUP is quietly unfolding. Its yardstick of value then shifts from PE to PS—from linear growth based on "earning money from each new store opened" to the exponential self-reinforcement of "data becomes more valuable with use, and models become more accurate with iteration." This fundamental change in asset attributes provides an irreversible industrial prerequisite for the valuation shift of RIMAG GROUP.

Rarity Realization: How the Full-Stack Technology Closed Loop Reshapes the Competitive Landscape of Medical Imaging AI

Possessing data infrastructure does not automatically equate to having AI capabilities. Transforming raw imaging data into trainable, verifiable, and iterable intelligent capabilities requires crossing three technical thresholds: data governance, foundational model, and productization closed loop. The technological path of Yinghe Medical Imaging has built systemic barriers at these three thresholds that are difficult for competitors to overcome.

The first threshold is data governance. The white paper explicitly states that data standardization is "an unavoidable prerequisite" for AI in medical imaging. Yinghe Medical Imaging's "Hetu Plan" is a systematic project addressing this prerequisite. As the "first and currently the only systemic data standard construction and data annotation plan in the medical imaging field," the Hetu Plan defines millions of medical imaging standardized semantics through standard infrastructure, creates precisely annotated datasets through data infrastructure, and supports general AI research and development through model infrastructure. This advances medical imaging AI from project-based development of "single-disease, single-model" systems to a large-scale development model centered around examination pathways.

The second threshold is the foundational model. Yinghe Medical Imaging's self-developed MIIA foundational model is trained from scratch as a medical imaging visual base model, giving it general representation, understanding, and generalization capabilities for CT, MR, X-ray, and other images. It then forms multimodal capabilities through image-text alignment. Unlike lightweight paths that directly call upon general large language models or external knowledge bases, the foundational model is the "underlying infrastructure for medical imaging AI," capable of "continuously absorbing new structured data and doctor feedback during real film reading, report revision, quality control feedback, and the accumulation of new cases, gradually adapting to the diagnostic habits of different hospitals, departments, and doctors."

The third threshold is the productization closed loop. For model capabilities to translate into clinical value, they must be embedded in real workflows and used continuously. The white paper clearly points out that one of the key pain points in the current industry is the "discrepancy between the training dataset and real clinical scenarios, which may lead to product performance in actual application falling short of R&D expectations." Yinghe Medical Imaging's MIIA product system builds a complete closed loop: "iSpaces generates data → iData governs data → iResearch trains models → MIIA-AI embeds into workflows → doctor feedback flows back," enabling AI capabilities to evolve continuously through use. Based on this closed loop, the cranial CT super-agent "Dr. Xiao Jun 2.0" can generate a complete structured report in less than one minute, improving report writing efficiency by approximately 1.5 times. The chest CT AIR achieves multi-disease identification within a single scan, with a comprehensive diagnostic accuracy of 84.6% and an AI report adoption rate of 86.7%.

The white paper positions the combination of RIMAG GROUP and Yinghe Medical Imaging as "the only full-stack closed-loop model globally." The cumulative effect of these three thresholds—data governance requires industry-level standard-setting capabilities and deep cooperation with authoritative medical institutions, the foundational model depends on continuous data supply, computing power investment, and algorithmic teams, and the productization closed loop necessitates real clinical scenarios and continuous doctor feedback channels—constitutes a systemic barrier. This is the underlying logic for the capital market to pay a scarcity premium for its industry position.

Valuation Rebuilding: From Traditional PE to Tech PS, a Strategic Shift in Market Cognitive Coordinates

Asset upgrade and rarity realization form the industrial prerequisites for valuation rebuilding, but the establishment of industrial logic does not automatically equate to the completion of capital pricing. The shift from PE to PS valuation requires the market to complete a re-cognition and confirmation of the company's industrial coordinates, its business closed loop, and its reference benchmarks across three levels.

The first level is the re-anchoring of industrial coordinates. The capital market's pricing of a company is essentially a pricing of its industrial position. RIMAG GROUP obtained its initial valuation anchor upon listing as "China's first listed company in third-party medical imaging services," based on Frost & Sullivan's previous industry position certification. Now, the white paper positions Yinghe Medical Imaging as a "representative technology platform in the field of medical imaging AI R&D and application," with its benchmark shifting from third-party imaging service institutions to leading tech platforms in the medical AI track—Yinghe Medical Imaging has built a "globally unique" full-stack closed-loop capability within the high-barrier, high-value vertical field of medical imaging. This shift in industrial coordinates provides a top-level cognitive framework for the valuation shift.

The shift in industrial coordinates needs to be realized through the verifiability of the business closed loop. The pricing shift in the capital market will not be driven by concepts alone; it requires quantifiable verification signals. On the data asset front, RIMAG GROUP pioneered the entire process exploration from "data resources to data assets to data transactions," with multiple imaging data products listed and traded on data exchanges, generating tiered data service revenue within "data sandboxes" of various state-owned platforms. On the AI product front, the cranial CT super-agent "Dr. Xiao Jun 2.0" and the chest CT AIR have already entered clinical use. The white paper's disclosed comprehensive diagnostic accuracy of 84.6% and AI report adoption rate of 86.7% provide quantifiable clinical validation of the product's value.

The continuous validation of the business closed loop will ultimately drive the market's shift in reference benchmarks. In 2025, TF Securities initiated coverage of RIMAG GROUP with a valuation benchmark against Tempus AI, signaling that the market has begun to revalue it as a "data asset operator" rather than an "imaging service provider." From the industry signal (Frost & Sullivan white paper) to business validation (data trading + AI product entry into hospitals) to valuation mapping (TF Securities' 8x PS initial coverage), all three levels of the coordinate system shift for RIMAG GROUP have been triggered. When the market confirms that a company simultaneously possesses the most scarce data infrastructure in the AGI era and the full-stack technical capability to convert data into iterable AI capabilities, the valuation leap from PE to PS is no longer a logical deduction but a pricing fact in the process of happening.

Summary

In conclusion, the valuation coordinate system of RIMAG GROUP is undergoing a strategic shift in its underlying logic reconstruction. On the asset level, the offline imaging network is being upgraded from "inspection capacity" to "AGI data infrastructure." On the technology level, Yinghe Medical Imaging has built a globally unique full-stack intelligent barrier through the Hetu Plan, the MIIA foundational model, and the "work-as-training" closed loop. On the valuation level, TF Securities' initial coverage at 8x PS marks the market's reference benchmark has already shifted from the traditional PE framework to the benchmarking system of medical AI data platforms. The white paper defines this combination as a new paradigm for the "digital intelligent imaging ecosystem." The combined value of RIMAG GROUP and Yinghe Medical Imaging jointly opens the "scenario-data-standard-model-product-feedback" digital intelligent imaging closed loop, "representing a new path for medical imaging AI to evolve from 'single-point algorithm products' to a 'digital intelligent imaging ecosystem.'" This is not only a certification of industrial position but also a significant signal for the reshaping of capital pricing logic. The cognitive shift in the capital market will take time, but the direction is clear: when a company simultaneously masters the most scarce data infrastructure of the AGI era and the full-stack technical ability to convert data into iterable AI capabilities, the market will ultimately measure its value with a new coordinate system.