Accelerating the Empowerment of the Health Industry with Large Models, China’s First Medical Corpus Terminal “Xiao Ku” Launched
The “Molding Shanghai, Future Health” forum on high-quality medical corpus applications for large models in healthcare, guided by the Shanghai Municipal Commission of Economy and Informatization, the Shanghai Municipal Health Commission, and the Shanghai Municipal Cyberspace Administration, was held in Shanghai on the afternoon of December 7. All images in this article are credited to Yu Kai, a reporter from The Paper.
To deeply implement the national strategy for developing a new generation of artificial intelligence, and to enforce the “Regulations on Promoting the Development of the Artificial Intelligence Industry in Shanghai” and the “Measures for Promoting the Innovative Development of Large Models in Artificial Intelligence (2023-2025),” the forum took place to address the challenges and needs in the application of medical large models. During the event, the “Application Guide for Medical Large Model Corpus Integration Machine” group standard was officially released, and the first medical corpus terminal in China, named “Xiao Ku,” was launched, along with a signing ceremony for a memorandum of cooperation on the construction of high-quality health corpus.
The forum focused on discussing the pain points and difficulties in the application of corpus in the field of medical large models, exploring the demands and models for their application, and releasing relevant standards and achievements to promote the development of medical large model corpus, thereby accelerating the empowerment of the healthcare industry by large models.
Yang Haijun, Chief Engineer of the Shanghai Municipal Cyberspace Administration, revealed in his speech that the administration, in collaboration with the Shanghai Municipal Health Commission and leveraging the expertise of the Pujiang Laboratory, is currently working on the establishment of a medical large model verification center. This initiative aims to explore a comprehensive evaluation of artificial intelligence applications in healthcare through a full-process review mechanism and professional scenario validation design, providing strong support for the compliant application of AI in the healthcare sector. He also mentioned that since last year, to promote the innovative development of the large model industry in Shanghai, the administration established the first compliance guidance service center for large models in the country and, in conjunction with the Shanghai Municipal Commission of Economy and Informatization, guided the establishment of the Shanghai Artificial Intelligence Security Governance Laboratory, as well as the formation of an AI Special Committee under the Shanghai Internet Association.
Zhang Hongtao, Deputy Director of the Shanghai Municipal Commission of Economy and Informatization, stated at the forum that in recent years, the artificial intelligence industry, represented by large models, has been accelerating its development, and Shanghai is striving to catch up with continuous innovation results. Currently, 60 large models have been filed in the city, rapidly being implemented in sectors such as healthcare, manufacturing, and education. The “Mosu Space” in Puxi and the “Moli Community” in Pudong are both vibrant, with an increasingly complete industrial ecosystem. Focusing on key industries such as healthcare, the commission is collaborating with renowned medical institutions like Renji Hospital to enhance scenario openness, conduct various supply-demand matching activities, and build demonstration bases for AI applications. Together with the China Academy of Information and Communications Technology Shanghai Innovation Center, Shanghai Kupar Technology Co., Ltd., and relevant medical institutions, they aim to create a batch of medical data resources for large models, accelerating the construction of typical and innovative demonstration application scenarios to serve as a model for the development of AI in healthcare nationwide.
Luo Meng, Deputy Director of the Shanghai Municipal Health Commission, pointed out at the forum that the “corpus” of large models can be broadly understood as the “data” people often refer to. High-quality health corpus is a necessary condition for enhancing the effectiveness of medical large models and expanding innovative application scenarios. Currently, medical large models still face challenges such as insufficient accuracy, lack of interpretability, and inadequate industry knowledge. Ensuring the quality and quantity of training data is key to translating the capabilities of large models into practical scenarios. Moving forward, the Shanghai Municipal Health Commission will collaborate with a wider range of innovative institutions to release and improve industry corpus, deepen data services and model innovation, and simultaneously build a corpus collection and governance system to promote the quantitative and qualitative enhancement of high-quality health corpus, providing solid support for the industry to embrace new large model technologies.
At the forum, Huang Haiqing, CEO of Kupar Technology, introduced the first medical corpus terminal product in China, “Xiao Ku.” The model area is equipped with nearly all the mainstream foundational large models, including “Shusheng,” “Qianwen,” and “SenseTime Rixin.” The corpus area contains industry-specific corpus databases, vector databases, and corpus processing tools, while the computing power area includes services from major computing power suppliers such as Birun, Cambrian, Huawei, and Suirun. “The birth of the corpus terminal aims to truly realize the vision of empowering various industries with large models,” said Huang Haiqing.
