Report: Health AI Index for China

Elsevier and Peking University Health Science Center publish report on AI in China from the perspective of research performance, knowledge transfer and media interest

By Wei Wang, PhD - January 22, 2021  7 mins
AI Health Index report main image
The Health AI Index report is available in Chinese. You can find the link below.

In China, an aging population, changing social burdens from chronic disease, and the aggravation and outbreak of new infectious viruses are creating significant challenges to public health. Addressing them requires an innovative and highly efficient mode of medical and healthcare services.

With its focus on the use of health artificial intelligence, China aims to leverage the power of AI and advanced science and technology to further its capabilities in medicine and health, build an extensive and optimized health ecosystem, and reduce variability in the provision of medical treatment.

To uncover insights in the area of AI-driven health sciences, Elsevier and the Peking University Health Science Center (PUHSC) have published the Health AI Index report (only available in Chinese).

PUHSC colleagues Elsevier worked with on the report.

Using the same data publication set as the 2018 Elsevier report Artificial intelligence: how knowledge is created, transferred and used, the Health AI Index looks at the scientific research performance, changes in academic research, experts’ perspectives and, for the first time, the R&D trend in the application of AI in health sciences through evidence-based analytics.

“The medical industry has entered the AI and big data era in a big way,” said Dr Erdan Dong, Head of the Department of Cardiology and Institute of Vascular Medicine at Peking University Third Hospital.

Erdan Dong, MD, PhD, Head of the Department of Cardiology and Institute of Vascular Medicine at Peking University Third Hospital, introduces the key findings of the Health AI Index report.

“Digital and intelligent medical care has become the research and development (R&D) standard in the health sciences, and China is one of the leading global AI players with considerable influence,” Dr Dong added. “PUHSC will continue to respond to the national call for the development of new AI technologies and, in the field of health sciences, support the development of AI and uncover the value of big data.”

Genan ZhaoGenan Zhao, Head of China Research Business at Elsevier, said Elsevier is committed to enabling the field of health sciences with quality data enriched by AI technology. He explained:

“Our collaboration with PUHSC on the Health AI Index report aims to uncover insights on the R&D trend of AI in the health segment through evidence-based analysis to provide a reference point for the strategic planning, R&D blueprint as well as clinical application on the management of health AI in China.”

Committed to focusing and tracking developments in the latest AI research and trends in the field of health, PUHSC created a composite index of scientific R&D accessible to global research institutions working in the field. Under the university’s leadership, the report was jointly developed by Elsevier and the following collaborators:

  • Health Big Data Application Management Professional Committee under the affiliation of China Hospital Association
  • PKU National Institute of Health Data Science
  • PKU Advanced Institute of Information Technology (Zhejiang)
  • PKU Center for Intelligent Public Health, Institute for Artificial Intelligence

Based on scientific publications and registered clinical trials, the report reviews and analyzes the scale, structure and development trend of global scientific research and clinical trials in health AI over a 5-year period from 2015 to 2019, with specific analysis on the performance of China.

Key findings of the report

Research performance

  • From 2015 to 2019, a total of 25,717 scientific publications related to the topic of health AI were published globally, and the normalized citation impact (FWCI) was 2.0. The top 5 countries with the highest scholarly output are the United States, China, India, the United Kingdom and Germany. Among Chinese institutions, the Chinese University of Hong Kong (CUHK) has the best comprehensive performance, ranking the top 10 in terms of the number of citations and FWCI.
  • In terms of research collaboration, China and the United States show a similar trend, which is reflected in the largest international collaborative article share; the least within institutions; the highest FWCI of international collaborative articles; and the lowest of intra institutional collaboration.
  • Although the academic-industry collaborative article share of China is lower than the world average, its research impact (FWCI = 3.7) is higher than the world average (FWCI = 3.4).
  • From a global perspective, the interdisciplinary fields of AI technology and health include medicine, computer science, physics, biochemistry, genetics and molecular biology, environmental science, engineering and social science. At present, algorithms, computer vision, models (in terms of the number of articles) and cytology; image segmentation; and medical imaging (in terms of the article share) have gained more attentions from academia. The research areas Chinese institutions are focused on also include algorithms, computer vision, and models. Magnetic resonance imaging, image segmentation, and medical imaging contribute the highest article shares.
  • Top 20 research topics with high research impact shows that health AI technology has been widely used in auxiliary detection, disease diagnosis and classification, genetics and genomics, large sample data prediction, etc. The diseases involved mainly include cancer and brain diseases, and the most important algorithm related research is convolution neural network.
  • Guangzhou Medical University, Sichuan University and Guangzhou Laboratory of regenerative medicine and health in Guangdong Province are in the top 20 worldwide institutions of high research impact in health AI. IEEE Transactions on Medical Imaging is the journal with the most published health AI-related content.

Knowledge transfer

  • Among the 44 articles cited by patents in China, the top five Chinese institutions are Chinese Academy of Sciences (CAS), Tsinghua University, Zhejiang University, Chinese University of Hong Kong and Shenzhen Institute of Advanced Technology. These patents focus on neural network, deep learning, heart arrhythmia, artificial intelligence and computer-aided diagnosis.

Media interest

  • In terms of health AI, mass media reports are different from the focus of S&T. Academia and industry focus on common diseases, frequently occurring diseases and public concerns in medical and health fields related to social facts.

Clinical practice

  • Since 2017, the number of clinical trials related to AI has increased rapidly globally, mainly from China and the United States. By September 2020, China has become the country having the largest number of AI related clinical trials in the world.
  • There is a lack of a universal design and evaluation standard for AI related clinical trials. At present, the research design specifications and report specifications of AI related clinical trials are still in the initial stage.
  • The main intervention measures of global AI related clinical trials are equipment intervention and diagnostic test intervention. The intervention measures of AI related clinical trials in China are mainly diagnostic test intervention and equipment intervention, while the United States mainly focuses on equipment intervention and behavior intervention.
  • Global AI related clinical trials mainly focus on people with cancer, major chronic diseases or psychological diseases. The outbreak of COVID-19 in 2019 has also attracted more attention.
  • At present, most of the clinical trials related to AI are still in the research stage, and the relevant test results have not been reported. In addition, there are still some limitations in the design of clinical trials related to AI.

Report summary

  1. The deep integration of advanced science and technology with medicine is the basis for the development of health AI, which will play a greater role in public health and clinical diagnosis in the future.
  2. China has become the main contributor of health AI scientific research and clinical trials, but its research impact and technology transformation still need to be improved.
  3. The core of health AI’s technology spectrum is in learning (including deep learning) and medical robots. The disease spectrum is mainly chronic diseases and nervous system diseases. There is still room for the deep integration of infectious diseases and rare diseases with medical AI.
  4. Compared with the hot research topics of the scientific community, the ethical issues of research and application of health AI have become the focus of social media.
  5. The global clinical research of health AI is still in the early stage. Universities/hospitals are the main sponsor in China, focusing on intelligent diagnosis of diseases; enterprise participation is still to be strengthened.

View the report

The report is free to read and download. It's available in Chinese.

AI Health Index report cover

Download the report


Related stories

AI main image
Clinical research image

Contributors


Wei Wang, PhD
Written by

Wei Wang, PhD

Written by

Wei Wang, PhD

Dr Wei Wang is a Collaboration Manager at Elsevier, based in Beijing. Her previous roles included Associate Professor at the Chinese Academy of Sciences. She received a PhD in Plant Sciences from the University of Alberta, Canada.

The guardians of Scopus
How collaboration changes the world for people with rare diseases
With a successful medical career, this researcher pursues his dream job

Comments


comments powered by Disqus