Yidu Tech Unveils New Product Yidu Evidence, Creating a "Evidence-Based Practice AI Assistant" for Doctors

2026-03-26

On March 10, Yidu Tech officially launched the clinical evidence-based agent "Yidu Evidence"—an "evidence-based assistant" designed specifically for doctors. It focuses on the real-world needs of doctors in clinical diagnosis, treatment, and research scenarios, breaking down the heavy barriers of traditional medical knowledge acquisition. This allows doctors to no longer be bogged down in literature review and evidence screening, enabling them to quickly obtain efficient decision-making support in clinical and research work, and to dedicate their valuable time to professional deepening and capability iteration, thereby driving the transformation and upgrading of clinical diagnosis and treatment models towards intelligence and evidence-based practice.

Yidu Evidence is now officially available for use. During the preceding one-week internal testing phase, Yidu Evidence attracted the active participation of over 6,000 medical professionals and received nearly 2,000 high-quality and valid feedback responses. The product's practicality and professionalism were highly recognized within the industry. Multiple clinical doctors who participated in the internal testing stated that Yidu Evidence performed excellently in tasks such as rapidly accessing guideline knowledge, obtaining treatment plans, querying drug interactions, assessing tumor staging, searching for clinical trials, and following cutting-edge advancements, significantly enhancing the efficiency of clinical information retrieval and doctors' capabilities for evidence-based decision support.

1. Rooted in Knowledge Evidence-Based Practice: More than a Tool, a Trusted "Evidence-Based Foundation"

The core value of genuine clinical-grade AI lies in professionalism and trustworthiness.

Yidu Evidence is built upon a rigorous evidence-based evaluation system, distilling a high-quality knowledge base from vast medical resources: over 30,000 authoritative institutional contents selected from more than 40,000 guidelines, and over 5 million high-quality scientific research results selected from over 30 million documents. Furthermore, in collaboration with authoritative institutions such as People's Medical Publishing House, it has constructed a disease and drug knowledge base reviewed by clinical experts, ensuring that every piece of knowledge originates from professional sources.

More importantly, Yidu Evidence is not a static knowledge base; it continuously tracks the latest research and findings from renowned journals such as NEJM, The Lancet, and JAMA, enabling doctors to access cutting-edge medical evidence in a timely manner, addressing the issue of outdated information common in traditional knowledge bases.

2. Every Statement Traceable, Making Evidence Levels Clear at a Glance

In clinical decision-making, doctors need not a "black box" that merely provides conclusions, but rather a "case file of evidence" that can withstand scrutiny. The core capability of Yidu Evidence lies in making the AI's generation process completely transparent, turning every piece of advice into verifiable and assessable evidence-based justification.

When a doctor poses a question to Yidu Evidence, the system does not simply retrieve a ready-made answer. Instead, it automatically executes a complete set of evidence retrieval, grading, and logical integration processes in the background. What is ultimately presented to the doctor is a clearly structured evidence-based response with rigorous annotation—

Every conclusion has a clear source. The system attaches the corresponding guideline or literature reference, allowing the doctor to verify the original material with one click. This not only builds the doctor's trust in the AI but also provides a traceable basis for medical quality and safety, clinical teaching, and difficult case discussions.

Every answer is sorted by evidence level. Based on evidence grading systems such as GRADE, the system automatically assesses the authority and timeliness of the evidence. When presenting the answer, it is strictly sorted according to the weight of the evidence level, freeing doctors from the effort of discerning the quality of information, making decision-making more efficient and precise.

3. Scenario-Based Implementation: Integrating AI into the Clinical Workflow

Clinical problems span multiple specialties and cover various scenarios, making it difficult for general-purpose AI to adapt to specific needs. Yidu Evidence features a built-in multi-specialty agent distribution engine that can intelligently identify the specialty attributes and clinical intent of a question, automatically routing it to the corresponding specialty agent. Each specialty agent is deeply trained using high-quality data within its domain, achieving "specialty-specific answers with precise adaptation," making the responses more relevant to clinical practice and more professional.

In addition to the general version, Yidu Evidence also offers a hospital-specific version. While strictly adhering to data security and compliance requirements, this version can be deeply integrated with a hospital's existing information systems. By incorporating individual patient diagnosis and treatment data, it can deliver more targeted and personalized evidence-based recommendations, truly achieving seamless integration between AI and the clinical workflow, and enabling the tangible implementation of intelligent capabilities.

Currently, Yidu Evidence has undergone rigorous clinical validation in hospitals such as the Sun Yat-sen University Cancer Center, covering 15 core tumor types including lung cancer, breast cancer, liver cancer, and colorectal cancer. The consistency rate between its treatment plan recommendations and clinical doctors' decisions is extremely high.

In an era of information explosion, the scarcest resource for doctors is not knowledge, but time. What Yidu Evidence does is not merely shorten the distance between a "question" and an "answer," but rather return precious time to doctors. Looking ahead, Yidu Tech will continue to drive product evolution through technological innovation, making Yidu Evidence a trusted intelligent companion by the side of every clinical practitioner, empowering the advancement of healthcare with the power of intelligent evidence-based practice.

本文为原创内容,仅供知识分享之目的,若存在侵权行为或疏漏,请与本平台联系,我们将及时处理。如需转载请后台留言,需征得本平台同意且注明来源转载。