In the current era of rapid advancement in artificial intelligence technology, the healthcare sector, a critical domain concerning life and health, is undergoing a profound paradigm shift from mere "technical prowess" to delivering "clinical value." Recently, the "Multimodal Intelligent Diagnosis and Treatment System Research and Transformation Application Beijing Key Laboratory," jointly established by Peking University First Hospital and YIDU TECH (02158), was officially inaugurated. This marks a significant step into a crucial phase of an initiative aimed at solving the "black box" problem in AI healthcare and creating a closed loop from data governance to clinical decision-making.
For a long time, AI's progress in high-risk medical scenarios has been hindered, with the core issue being the trust gap created by its "unexplainable" nature. What doctors need is not just a simple diagnostic conclusion, but a decision-making path that can withstand scrutiny. YIDU TECH's solution centers on the principles of "evidence-based and traceable." Leveraging its self-developed "AI Medical Brain," YiduCore, the company provides the laboratory with a robust data governance foundation, transforming multi-source, heterogeneous data into high-quality assets.
Building on this, YIDU TECH has constructed the "Yidu Zhi Xun" evidence-based intelligent agent platform. This platform integrates large language model semantic understanding with medical knowledge graphs to achieve integrated decision support encompassing "problem understanding—evidence retrieval—reasoning generation—result tracing." When doctors receive AI recommendations, they can clearly see the supporting guidelines and literature evidence behind them. Currently, the clinical problem response time is under 20 seconds, evidence retrieval efficiency is improved by 70% to 90%, and the source traceability rate is close to 100%.
The establishment of the "Multimodal Intelligent Diagnosis and Treatment" key laboratory is a microcosm of YIDU TECH's deep integration into Beijing's "AI + Healthcare" strategy. The laboratory's first major target is cardiorenal metabolic syndrome (CKM), aiming to break down specialty barriers and promote a shift towards intelligent, whole-process management for "preventing disease and treating comorbidities." This model of deep integration between industry, academia, research, and medicine is transforming the "evidence-based" concept into a replicable practical pathway.
Presently, YIDU TECH has collaborated with leading hospitals to develop over 280 specialized scenario intelligent agents, deeply embedded into real clinical environments like doctor workstations. These are deployed in 133 renowned hospitals, including Peking Union Medical College Hospital, Peking University Cancer Hospital, and Tsinghua Chang Gung Hospital, covering high-frequency, full-process needs such as medical record generation, assisted diagnosis, and risk assessment.
Simultaneously, YIDU TECH is accelerating its layout at a broader ecosystem level. At the 2026 Global Digital Economy Conference, as a founding member, YIDU TECH participated in launching the "Chaoyang District Healthcare Data Element Ecosystem Alliance," aimed at connecting the entire chain of healthcare data from rights confirmation and circulation to value realization.
At the National AI Application Pilot Base (Healthcare), YIDU TECH showcased a patient recruitment intelligent agent that has increased AI screening accuracy for clinical trials to 80%-90%, while reducing manual screening costs by nearly 90%.
The ceiling for AI in healthcare is not determined by model parameters, but by high-quality data infrastructure and a deep understanding of real clinical scenarios. Only when "evidence-based" becomes a standard feature of intelligent agents can AI truly qualify to enter the life-critical decision-making loop. The realization of this technological deployment is reflected in the latest financial report.
The fiscal year 2026 results disclosed on June 30th show that YIDU TECH achieved its first full-year profit since listing, with an annual profit of 78.766 million yuan, exceeding the upper limit of the April profit alert by approximately 12.5%. Net profit attributable to owners was 72.709 million yuan, 30% higher than the upper limit of the forecast. Operating cash flow turned positive in the second half to 29.3 million yuan, and the board recommended its first dividend of HK$0.04 per share, verifying the sustainability of profitability.
As the medical AI industry moves from "competing on parameters" to the second half of "competing on trustworthiness and implementation," the scarce capabilities embedded in the entire chain of regulation, clinical practice, and data are precisely the firmest anchors for valuation.
Comments