China Launches Clinical Evidence-Based AI Capability Initiative with Yidu ZhiXun Debut at Zhongguancun Forum

On March 26, at the “World Digital Health Forum” of the 2026 Zhongguancun Forum Annual Conference, a batch of “Medical Technology Innovation Achievements” was released, and the “China Clinical Evidence-Based Intelligence Capacity Building Initiative” was officially launched as one of its major outcomes. This is the first industry-level collaborative initiative in China to focus on the deep integration of evidence-based medicine and artificial intelligence. The initiative was jointly initiated by several renowned medical institutions in China, including Beijing Tsinghua Changgung Hospital of Tsinghua University, with Yidu Tech, a leading domestic AI healthcare enterprise, providing core AI capabilities. At the same time, the platform’s core technological vehicle—the clinical evidence-based intelligent agent platform “Yidu Zhixun” APP—was launched and made available to physicians.

Currently, a new medical paper is published every 30 seconds worldwide, and clinical guidelines are updated frequently. However, the efficiency with which frontline clinicians access and apply high-quality evidence-based medicine remains insufficient—a contradiction that is particularly pronounced within the tiered diagnosis and treatment system. The newly launched “China Clinical Evidence-Based Intelligence Capacity Building Initiative” is designed precisely to address this structural pain point.

Leveraging a multi-center collaborative mechanism involving clinical experts, this initiative systematically integrates real-world medical evidence, domestic and international clinical guidelines, and cutting-edge research findings. Its goal is to construct an evidence-based intelligent capability framework that aligns with the realities of clinical practice in China and is scalable for widespread adoption. Medical institutions, including Beijing Tsinghua Changgung Hospital affiliated with Tsinghua University, are deeply involved, establishing a closed-loop co-construction model that spans from specialty knowledge validation to adaptation in clinical scenarios.

As a core participant in this initiative, Yidu Tech is responsible for underlying knowledge modeling and the delivery of AI capabilities, with “Yidu Zhixun” serving as the core technological vehicle and commercialized product. Previously, Yidu Tech’s AI solutions have undergone extensive clinical practice at institutions such as Peking Union Medical College Hospital, Sun Yat-sen University Cancer Center, and Peking University Cancer Hospital, accumulating end-to-end expertise ranging from data governance to agent construction. Leveraging this industry collaboration initiative, Yidu Tech will extend its intelligent evidence-based capabilities to broader scenarios, enabling more physicians to conveniently access high-quality evidence-based support.

Xu Jiming, Co-founder and CEO of Yidu Tech, stated that Yidu Zhixun transcends the limitations of traditional “information retrieval tools.” With a clear positioning as a “super assistant, second brain, and multidisciplinary AI expert advisory board,” it aims to become the core support for physicians’ clinical work.

For physicians, Yidu Smart Guidelines serves as a “super assistant” that balances efficiency with personalization. By constructing an evidence-based knowledge base covering over 30,000 authoritative clinical guidelines and more than 5 million curated medical literature articles, Yidu Smart Guidelines has established a standardized knowledge system in collaboration with institutions such as the People’s Medical Publishing House, while continuously tracking updates from leading global journals. Whether for prescribing during outpatient visits, making decisions during ward rounds, or discussing cases, doctors can obtain precise conclusions within seconds simply by inputting their clinical questions.

 Yidu Zhixun App Function Demonstration

Even more thoughtfully, the “Personal Knowledge Base” feature functions as an exclusive “mobile study,” enabling physicians to bookmark and upload approved guidelines and literature, categorize them according to personal preferences, and pose queries, thereby creating a personalized knowledge space.

Clinical decision-making allows for no ambiguity. Adhering to the principle of "traceability for every statement," Yidu Zhixun equips each conclusion with an "evidence ID," enabling one-click retrieval of the original source. Meanwhile, it grades the quality of retrieved evidence according to the international GRADE system and incorporates multiple validation mechanisms to automatically verify literature support, eliminating unsubstantiated conclusions. This systematically reduces the risk of "AI hallucinations" at the technical level.

Yidu Zhixun App Function Demonstration

This fully transparent and verifiable decision-making logic serves as a “second brain” for physicians, compensating for the limitations of human memory while providing rigorous evidence-based support, thereby making every clinical decision more secure and reliable.

The multi-specialty and multi-scenario nature of clinical diagnosis and treatment makes it difficult for general-purpose AI to meet the needs of specialized fields. Yidu Zhixun has innovatively built a “Multi-Specialty Agent Collaboration Engine,” which functions like assembling N “expert advisory teams” for physicians. The system can automatically identify the specialty attributes and clinical intent of inquiries, and proactively and precisely route them to the corresponding specialty-specific agents for processing.

Yidu Multi-Agent Engine

These agents have been meticulously refined within hospital settings: Yidu Tech has developed over 200 specialty-specific agents, deployed and validated in real-world clinical environments across multiple partner hospitals. Trained on domain-specific data and rigorously verified in clinical scenarios, these agents address the limitation of general-purpose AI—“broad but not deep”—by providing decision support that aligns more closely with actual diagnostic and treatment workflows.

Addressing the in-depth needs of healthcare institutions, Yidu Smart Evidence simultaneously launches a hospital-exclusive version. This version allows for flexible adjustment of specialist agent weights based on the hospital’s disciplinary specialties, precisely aligns with the institution’s drug formulary, clinical practice guidelines, and quality control requirements, and supports the co-construction of exclusive knowledge bases with in-house experts. More critically, this version can be deeply integrated with the hospital’s existing information systems, such as Hospital Information Systems (HIS) and Electronic Medical Records (EMR), enabling seamless linkage between patient data and evidence-based knowledge bases, thereby truly embedding intelligent evidence-based decision support into clinical workflows.

From “information tool” to “gateway for decision-making,” from “generalized AI” to “specialist-level intelligence,” and from “passive inquiry” to “proactive integration into workflows”—Yidu Tech’s Yidu Zhixun is redefining the value boundaries of AI in healthcare. Launched at the Zhongguancun Forum, the “China Clinical Evidence-Based Intelligence Capacity Building Initiative” marks not only the starting point for industry collaboration but also a significant milestone in advancing China’s “evidence-driven decision-making” in healthcare toward greater intelligence and accessibility.

"Ground clinical decision-making in evidence, ensuring every statement is supported by data. Yidu Smart Evidence is turning this vision into reality."

Click to download and experience the Yidu SmartCycle APP