Global First! Chongqing Medical University Affiliated Second Hospital Launches AI Assistant for Hepatitis B Cure

Recently, Yidu Tech and the Second Affiliated Hospital of Chongqing Medical University (hereinafter referred to as “the Second Affiliated Hospital of CQMU”) jointly createdThe World's First "AI Assistant for Hepatitis B Cure"Officially released at the 17th Academic Conference on Challenges and Hotspots in Antiviral Therapy for Chronic Viral Hepatitis.

This innovative application is built upon Yidu Tech’s proprietary large language model for the medical vertical, providing patients with a one-stop, closed-loop service that spans “screening, diagnosis, treatment, follow-up, and research.”

The AI Assistant Released This TimeDeeply integrated with clinical pathways,The system aims to address issues such as fragmented information, inefficient clinical trial recruitment, and gaps in management within the diagnosis and treatment of hepatitis B, providing patients with full-cycle, closed-loop services and serving as an “intelligent partner” integrated into the entire care process.

The AssistantBuilt on the “Hepatitis B Cure Full-Cycle Intelligent Management System” jointly developed by both parties, and powered by a large medical language model integrated with Retrieval-Augmented Generation (RAG) technology, it establishes a digital support system covering the entire disease course.

A knowledge base was constructed based on an authoritative knowledge system comprising professional guidelines, literature, and institutional protocols within the field. This knowledge base underwent systematic analysis and structured processing, enabling AI to accurately comprehend medical concepts and effectively correlate diagnostic and therapeutic logic, thereby providing users with evidence-based decision support.

■ In patient management, the AI assistant for hepatitis B cure acts as a "health steward", enabling generative follow-up, personalized education, health consultations, and appointment reminders for cohorts of hepatitis B patients, breaking down barriers between hospital departments and institutions to provide patients with uninterrupted intelligent services.

■ In terms of diagnostic and treatment collaboration, it serves as an efficient “clinical collaborator”, the constructed full-cycle closed-loop intelligent management system covers the entire process of high-risk population screening, diagnostic assistance, personalized treatment, long-term follow-up, and rehabilitation guidance, while also breaking down data barriers between scientific research and clinical practice to facilitate the rapid translation of research achievements.

The precision and reliability behind the AI assistant stem from the long-term collaboration between Yidu Tech and the Second Affiliated Hospital of Chongqing Medical University in data infrastructure and scenario deepening. Previously,The two parties have jointly established a hospital-wide scientific research data platform and, with a focus on liver diseases, have collaboratively built a high-quality “Hepatitis B Specialized Disease Database.”

Building on this foundation, and leveraging Yidu’s “AI Middle Platform,” the team conducted specialized training and scenario-based fine-tuning of large language models tailored to the medical vertical, while integrating a Retrieval-Augmented Generation (RAG) mechanism to ensure the accuracy and timeliness of information output. This enables“AI Assistant for Hepatitis B Cure” not only answers frequently asked questions but also comprehends complex medical logic, providing personalized risk assessment and decision support.

The efficacy and professionalism of this capability have been validated in real-world clinical settings. Previously, Yidu Tech launched its independently developed and controllable “Clinical Copilot,” which leverages an integrated architecture of “big data + large language models + AI agents” to provide lightweight, all-in-one assistance across the entire diagnosis and treatment workflow.

It seamlessly integrates into physician workstations, enabling doctors to autonomously configure AI assistants using “Lego-style” modular components to achieve “co-creation of AI capabilities.” The solution has demonstrated high adoption rates and practical value in numerous leading hospitals and has been recognized by industry media as “one of the most promising Chinese AI products poised to catch up with OpenEvidence.”