YiduTech's AI Platform Successfully Deployed at Xiangya Hospital of Central South University: China's First Deep Integration with DeepSeek AI Middleware

On February 20, it was reported that Xiangya Hospital of Central South University had successfully completed the localized deployment of a domestic AI healthcare middle platform developed by Yidu Cloud on the 18th. This platform supports the joint localized deployment, invocation, and training of domestic large language models such as DeepSeek, providing a new paradigm for the construction of smart hospitals.

This successful deployment marks Xiangya Hospital of Central South University as the first institution in China to deeply integrate DeepSeek into its AI middle platform. The Information Center will collaborate with the hospital’s physicians to develop agent-based applications leveraging the institution’s years of accumulated clinical data., jointly promoting the enhancement of the hospital's AI-driven new quality productive forces.

 Yidu Cloud’s AI Middle Platform Empowers Hospitals’ Smart Transformation from Infrastructure to Applications

It is understood that the deployment at Xiangya Hospital of Central South University this timeThe innovation of the AI Middle Platform lies in its “Four Opennesses,” namely openness in technology, data, services, and ecosystem:

● Technology OpennessSpecifically, based on the middle-platform technical specifications, it is possible to integrate with domestic large language models such as DeepSeek-R1’s universal large model and Yidu Cloud’s medical vertical-domain large model, supporting rapid expansion and iteration of multi-dimensional large models.

● Data OpennessNamely, the ingestion, governance, and sharing of multidimensional data.

● Service OpenIt provides a variety of pre-trained models and algorithms, and supports user customization.

● Open EcosystemIt enables the free development, training, and deployment of large model applications through agents. Meanwhile, the AI Open Middle Platform, built on Ascend GPUs, achieves end-to-end autonomy and controllability from hardware to software, significantly enhancing security.

Xiangya Hospital of Central South University has always been at the forefront of the industry in the construction of smart hospitals,Aggregating over 2.5 billion medical data records and processing them through medical logic, we have established more than ten specialized disease databases, building a highly influential data center in China. The hospital has comprehensive reserves in both software and hardware capabilities, creating a solid and efficient “digital foundation” for the integration of large language models. This ensures more stable and seamless data utilization, providing robust support for subsequent complex medical data processing and analysis tasks.

Yidu Cloud, a leading AI healthcare company in China, has been deepening its collaboration with Xiangya Hospital of Central South University in the field of smart healthcare since 2019.Established China’s first artificial joint registry system, a big data platform for gynecologic oncology, specialized research databases for cerebrovascular diseases, perioperative care, and sepsis, as well as hospital-wide and multi-center research data platforms, and an intelligent data acquisition platform for severe COVID-19 research.

Previously, Yidu Cloud announced that it had integrated the DeepSeek artificial intelligence model into its independently developed “AI Healthcare Brain,” YiduCore, further driving the large-scale application and innovative practice of AI technology in the healthcare industry.. The deployment of the AI middle platform at Xiangya Hospital of Central South University is also a benchmark case for Yidu Cloud’s rapid application implementation following its integration with DeepSeek.

According to public information from Xiangya Hospital of Central South University, the AI middle platform will integrate with more business systems to expand into a wider range of rich application scenarios. For instance, in clinical decision support, large models will combine multidimensional data—such as patient medical history, symptoms, and test results—to provide diagnostic recommendations and treatment plan references for physicians. In medical record quality control, large models help improve the quality of medical records. Furthermore, large models can accurately identify pathological features in medical images, assisting doctors in detecting diseases at an earlier stage. Large models will also play a significant role in daily administrative tasks to enhance work efficiency; for example, by leveraging large models to conduct formal and content reviews of research grant applications, researchers can focus on designing key innovative components.

Xu Jiming, Co-Founder and CEO of Yidu Cloud, stated that 2025 will be a breakout year for the industrial application of large language models.Yidu Cloud is accelerating the deep integration of its self-developed large language model for the medical vertical into clinical diagnosis and treatment, hospital management, scientific research, and other scenarios through technological synergy with the DeepSeek large language model.Build an intelligent ecosystem covering the entire medical field to promote the digital transformation of the healthcare industry.