At the recent 2025 Cancer Prevention and Treatment Progress Conference in Beijing, JD HEALTH (06618) announced significant advancements in its "Jingyi Qianxun" medical AI model. The platform has evolved from a general-purpose base model to a comprehensive medical AI system, now focusing on disease-specific applications. The company is collaborating with top-tier medical institutions in China to develop specialized models covering common cancers and major chronic diseases.
During the conference, JD HEALTH partnered with the Chinese Anti-Cancer Association to launch the "Integrated Cancer Prevention and Treatment Project (2025-2030)." This initiative aims to establish a full-cycle system spanning prevention, screening, diagnosis, treatment, and rehabilitation, integrating cutting-edge AI technologies into clinical practice to explore China's unique approach to cancer care. The company also signed AI research agreements with leading hospitals including Guangzhou Medical University Affiliated First Hospital, Sun Yat-sen Memorial Hospital, and Peking University Cancer Hospital, driving systemic innovation in oncology diagnostics and treatment.
**Deep Specialization: Setting New Benchmarks in Disease-Specific AI** Since 2018, JD HEALTH's medical AI has progressed through three key phases: from "knowledge graphs + traditional small models" to "disease diagnosis pathways + billion-parameter models," and now to "health knowledge base + trillion-parameter models." In 2025, the company has focused on refining its "Jingyi Qianxun" foundation model for specialized medical fields while promoting open-source collaboration to address healthcare accessibility challenges.
The model leverages high-quality online consultation data from JD Internet Hospital and an evidence-based repository of clinical guidelines and research papers. Through partnerships with national medical data centers, it utilizes de-identified large-scale medical records and imaging data to strengthen its foundation. Collaborations with top hospitals further integrate multidimensional clinical data—including medical history, lab tests, imaging, and pathology—for specialized model training and real-world validation.
In oncology, the disease-specific models employ a dual approach of "real-world alignment + expert knowledge integration." By processing comprehensive clinical data and incorporating insights from over 1,000 specialists (including nearly 100 oncology experts) across 400+ tertiary hospitals, the AI replicates individualized diagnostic reasoning, creating clinically aligned "digital counterparts." This ensures both precision and broad dissemination of expert-level medical intelligence.
Current applications include: - **Gastrointestinal cancers**: A model co-developed with Peking University Cancer Hospital covers esophageal, gastric, and colorectal cancers. - **Respiratory diseases**: A collaboration with Guangzhou Medical University supports lung cancer, COPD, and pneumonia management. - **Urologic tumors**: Breakthroughs with Sun Yat-sen Memorial Hospital enhance precision diagnostics for prostate cancer.
These models enable end-to-end clinical reasoning—from diagnosis to imaging analysis and treatment planning—embedding AI across every critical phase of cancer care.
**Ecosystem Enablement: Building Trusted AI Collaborators** JD HEALTH is deploying its specialized models to create an intelligent cancer management system spanning prevention to rehabilitation, positioning AI as a reliable clinical partner. Key implementations include: - **Clinical services**: JD Internet Hospital's oncology department engages 30,000+ specialists, using AI assistants to automate 90% of routine queries and improve consultation efficiency by 50%. Its industry-first "open-dialogue system" enables multimodal interactions and 24/7 expert-level virtual consultations. - **Research acceleration**: An AI-powered research platform streamlines data collection and follow-ups, exemplified by an ongoing 20,000-patient NSCLC study using decentralized clinical trial (DCT) methods to enhance data quality. - **Patient management**: AI enables granular, full-cycle tumor monitoring via smart follow-ups, rehabilitation guidance, and streamlined re-examinations, improving compliance and reducing redundant consultations. Integrated health records further personalize long-term care strategies.
Looking ahead, JD HEALTH will advance its models through cognitive闭环 systems, cross-modal upgrades, and deeper clinical reasoning. Committed to open collaboration, the company aims to build an accessible, evidence-based, and patient-centric specialty care ecosystem—accelerating the realization of "Healthy China 2030" through technological innovation.
Comments