China New Growth: Large language models lending hand in healthcare scenarios

BEIJING, Sept. 24 (Xinhua) -- China's fast-growing AI sector has picked up pace in addressing some long-standing headaches facing human beings. This sector is making a contribution by leveraging large language models to offer smart medical services and facilitate developing medicines.

Following the launch of its AI chatbot Ernie, Chinese AI giant Baidu last week rolled out an industry-grade medical large language model called Lingyi, which can analyze and generate patient data to aid diagnosis and treatment and serve as a 24-hour "healthcare manager" for patients.

The company said that the large language model will be available for trial use in both upstream and downstream healthcare sectors, in a bid to improve the digitization and intelligence of the healthcare industry.

Several online medical services platforms have also made forays into AI-empowered large language models.

In May this year, digital healthcare service provider Medlinker unveiled MedGPT that seeks to generate practical diagnostic value in real medical scenarios. In June, counterpart Dingdang Health launched Dingdang HealthGPT and introduced application-based smart pharmacist and nutritionist assistant AI products.

"Modern medicine, life science, and engineering science are highly integrated, which promotes the evolution of a medical model from engineering medicine to intelligent medicine," said Jin Li, academician of the Chinese Academy of Sciences and director of Shanghai Medical College of Fudan University, while also adding that AI plays a key leading role in this process.

In addition to daily-life medical services, some tech firms have sought to boost pharmaceutical effectiveness and efficiency through AI.

Pharmolix, a biomedical start-up incubated by the Institute for AI Industry Research (AIR), Tsinghua University, has developed Chat Drug Design (ChatDD), a computing product that seeks to assist and promote conversational biomedical research and development.

The ChatDD effectively links the knowledge of experts and large language models through human-machine collaborative dialogue, pioneering a new path for drug R&D, said AIR's dean Zhang Yaqin.

Drug development is a field that can deeply integrate many academic articles and data to generate good outcomes, as the large language model has incorporated more nuanced and sophisticated data and can offer molecular design optimization, according to Nie Zaiqing, Pharmolix's chief scientist.

Some governments have sought to keep pace with latest innovations in the healthcare sector.

Home to two zones of pharma and life science innovation, Shanghai has been a pioneer among Chinese provincial-level regions in the exploration of deep integration of AI and healthcare.

Shanghai will develop AI algorithms and large language models suitable for life science problems, provide efficient, accurate and interpretable algorithm services, and promote innovative applications in medical big data, medical imaging, gene sequencing and drug R&D, according to Zhang Ying, deputy director with the municipal commission of economy and information.

In terms of medical AI, China has carried out applications in clinical assisted diagnosis and treatment, public health intelligent services, and medical intelligent management, and achieved concrete results, according to Shen Jianfeng, an official with the National Health Commission (NHC).

The NHC will improve the mechanism design related to medical AI services and explore their promotion by strengthening sci-tech collaborative innovation and encouraging cross-industry integration to achieve a well-regulated development, Shen said.