On December 27, Tencent announced the launch of the world’s first serious illness medical model, the Qiyuan Critical Care Model, developed in collaboration with Mindray Medical. This groundbreaking technology aims to address the challenges faced in Intensive Care Units (ICUs) around the globe.
The Qiyuan Critical Care Model boasts an impressive trillion-parameter count and has been trained on extensive medical text data. It encompasses 2.85 million medical entities and 12.5 million medical relationships, creating a comprehensive medical knowledge graph and literature that covers 98% of medical knowledge.
Currently, the model is being piloted in the ICU of the First Affiliated Hospital of Zhejiang University School of Medicine. It offers real-time monitoring of patients’ vital signs and provides 24-hour assistance to doctors, enabling faster treatment and diagnosis for critically ill patients. This allows healthcare professionals to focus their time and efforts on patient care.
Within just five seconds, the model can generate a summary of a patient’s condition by analyzing their vital sign data and establishing a digital profile. Doctors no longer need to sift through multiple information systems; they can simply query the Qiyuan model for an accurate patient history.
Once the digital profile is created, the model conducts a thorough analysis using the mindset of critical care physicians, offering diagnostic and treatment recommendations to assist healthcare providers with subsequent intervention efforts.
In addition to its analytical capabilities, the Qiyuan model excels in understanding and generating medical texts, accurately interpreting medical inquiries. Clinical simulations indicate that the model’s recommendations and critical care analyses have an impressive accuracy rate of 95%.
Furthermore, the system can automatically and efficiently generate patient medical records based on the existing clinical data. The produced documentation is coherent, conforms to clinical standards, and requires minimal adjustments by doctors, enabling them to complete what used to be a 30-minute documentation task in just one minute—boosting efficiency by 30 times.
This innovative approach has the potential to transform critical care practices, significantly enhancing both the speed and accuracy of patient assessments and record-keeping in medical settings.