Significant progress has been made in the field of Chinese medicine. Recently, the "Xiehe Taichu" rare disease AI large model, jointly developed by Peking Union Medical College Hospital and the Institute of Automation, Chinese Academy of Sciences, has officially entered the clinical application stage. This not only marks a breakthrough in China's rare disease diagnostic technology but also provides a new tool for improving clinical efficiency.
The development of this AI large model benefits from years of accumulation of rare disease knowledge bases in China and the support of Chinese population genetic testing data. As the world's first rare disease large model adapted to the characteristics of the Chinese population, its purpose is to help doctors identify and diagnose rare diseases more quickly and accurately, significantly shortening the time to diagnosis for patients. This is undoubtedly a major boon for rare disease patients.
Traditional artificial intelligence model development often requires massive amounts of data for training. However, due to the scattered nature of rare disease cases and the scarcity of data, traditional methods are difficult to apply effectively. To address this, the research team adopted an innovative minimal sample cold-start technology, requiring only a small amount of data combined with medical knowledge to achieve auxiliary decision-making functions throughout the entire diagnostic and treatment process. This innovation makes the application of the AI large model in clinical settings more feasible.
Furthermore, the model successfully incorporates the deep reasoning capabilities of DeepSeek-R1, possessing three core advantages: decision-making logic conforms to clinical thinking patterns, effectively suppressing AI hallucinations, and supporting autonomous knowledge iteration. These advantages enable the model to better adapt to the complex and ever-changing needs of clinical practice, providing strong support for doctors' decision-making.
Currently, the initial consultation and appointment functions of this AI large model are open for patient testing. In the future, it will gradually be integrated into the online diagnosis and treatment services of Peking Union Medical College Hospital's rare disease joint outpatient clinic, and plans are underway for nationwide promotion to further promote the construction of China's rare disease diagnosis and treatment collaboration network.
The successful application of this technology not only showcases China's cutting-edge technological level in medical artificial intelligence but also points the way for the future diagnosis and treatment of rare diseases, greatly improving patient medical experience and efficiency.