《CSET：2024中國醫療AI：生物大數據對生物經濟的影響研究報告（英文版）（36頁）.pdf》由會員分享，可在線閱讀，更多相關《CSET：2024中國醫療AI：生物大數據對生物經濟的影響研究報告（英文版）（36頁）.pdf（36頁珍藏版）》請在三個皮匠報告上搜索。

1、Issue BriefMay 2024China and Medical AIImplications of Big Biodata for the BioeconomyAuthorsCaroline SchuergerVikram VenkatramKatherine QuinnCenter for Security and Emerging Technology|1 Executive Summary Medical artificial intelligence is an area of emerging technology that leverages biomedical adv

2、ances and AI to power new discoveries for human health.Medical AI can help researchers discover new medicines,understand complex causes of disease,and predict how patients will respond to therapies.As these applications improve public health outcomes,they also contribute to the growing global bioeco

3、nomy.Countries that strategically prioritize medical AI could enjoy a competitive economic advantage and set global standards and norms for future developments.Medical AI leadership depends on access to large repositories of biological data,or biodata.Biodata can include information collected from i

4、ndividuals,such as their unique genetic sequences or medical records.Researchers and developers who have access to large amounts of high-quality biodata are at an advantage,as this resource is a limiting factor for medical AI advancement.China has created a comprehensive national strategy to support

5、 medical AI development and advance its goals for bioeconomy leadership.This report examines Chinas stated goals for medical AI,which range from the collection and protection of vast amounts of biodata,to facilitating research and development,to supporting medical AI commercialization.We find that:C

6、hina has access to publicly-available biodata from around the world,while its domestic datasets are closed off to other countries.Access to more biodata facilitates more powerful medical AI applications.Medical AI research publications from both China and the United States are on the rise.The United

7、 States only narrowly led China in the production of research publications related to medical AI in 2021,and Chinas research output is likely to continue to grow due to the many resources the Chinese government is pouring into this area.Beijing is promoting medical AI commercialization,and top Chine

8、se technology companies that have not previously been involved in biotechnology are moving into the medical AI space.Policies to support companies and facilitate regulatory approval can accelerate medical device development.In sum,Chinas strategy for biodata collection and medical AI development cou

9、ld position it to be an economic and technological leader in this sector.Policymakers should consider the implications of this leadership when considering policies to boost U.S.competitiveness and biodata infrastructure.Center for Security and Emerging Technology|2 Table of Contents Executive Summar

10、y.1 Introduction.3 Biological Data:A Societal Resource Fueling AI Innovation.4 Policies to Make China a Global Biodata Leader.9 Policies Facilitating Biodata Accumulation.9 Policies Controlling Access to Data.11 Policies to Support Research,Development,and Technology Infrastructure.13 Chinas Medical

11、 AI Research Output.13 Chinese Companies Developing Biodata-Enabled Medical AI.16 Conclusions and Lasting Implications of Chinas Biodata Leadership.19 Appendix A:Select List of Chinas Notable Governmental Policies on Medical AI.20 Authors.24 Acknowledgments.24 Endnotes.25 Center for Security and Eme

12、rging Technology|3 Introduction As science and technology advance,cutting-edge developments in artificial intelligence are accelerating biological innovation.AI allows researchers to process enormous amounts of complex data more quickly and efficiently,and can advance health and medical applications

13、 by identifying patterns in large repositories of biological data(hereinafter referred to as biodata).For instance,during the COVID-19 pandemic,countries around the world used AI tools to diagnose disease,predict therapeutic responses,and calculate infection probabilities.1 Looking ahead,there will

14、be huge economic and societal rewards for countries that lead the development and deployment of AI technologies for human health purposes(hereinafter referred to as medical AI).The global bioeconomy is rapidly growing;some experts estimate that bio-produced products could contribute up to$4 trillion

15、 per year to the global economy by the end of this decade,and that medical AI alone will be worth over$12 billion globally by 2030.2 In addition to economic power,the leader in medical AI can set global standards for research,technology deployment,and data usage.Given these potential advantages,coun

16、tries have an incentive to invest resources and create dedicated policies to advance the technological development of medical AI,including by making high-quality biodata repositories more accessible.3 China is one of the best examples of a country that has created a strategy to integrate biotechnolo

17、gy and AI.China can access publicly-available data from around the world,while maintaining domestic datasets that are closed off to other countries.4 Many Chinese government policies explicitly state their intent to prioritize biodata collection,application,and medical AI deployment,as do statements

18、 from Chinas medical AI industry.5 While we analyze those published declarations and strategies,their implementation and eventual success remain to be seen.This brief examines Chinas stated and understood goals for biodata and medical AI in a global context,focusing on the policies and programs the

19、Chinese government has put in place to collect and analyze vast amounts of biodata including genetic,proteomic,and clinical data.It also examines different aspects of the ecosystem that China is developing for cutting-edge technologies,including medical AI,in order to compete and dominate this growi

20、ng aspect of the bioeconomy.This brief concludes by assessing the implications of these developments for U.S.competitiveness in this sector.It is important to note that,while this report acknowledges the inherent national security and ethical implications associated with Chinas efforts surrounding b

21、iodata,it focuses primarily on the economic implications of these efforts.Center for Security and Emerging Technology|4 Biological Data:A Societal Resource Fueling AI Innovation Biological data,or biodata,is a broad category that encompasses large datasets of biological,life sciences,or biomedical i

22、nformation.Among other examples,biodata can include information collected from individuals,such as their unique genetic sequences or medical records(types of biodata summarized in Table 1).AI technologies can make use of this huge amount of data to identify links between genes,diseases,and treatment

23、s that would otherwise be difficult and time-consuming to recognize.Cutting-edge medical AI applications are increasingly being adopted in fields like personalized medicine,drug development,and disease surveillance.As these tools are implemented,their use generates and collects even more biodata tha

24、t can feed into the next cycle of development(Figure 1).Center for Security and Emerging Technology|5 Figure 1.Cycle of AI-Enabled Medical Technology Development Source:CSET analysis.Center for Security and Emerging Technology|6 Increased access to biodata can support and accelerate basic research,a

25、nd help scientists and doctors analyze demographic health trends or such trends across the general population.6 That said,making efficient use of biodata and particularly DNA data resources can be challenging and time consuming.This is in part because most biodata datasets are large;for example,ther

26、e are about three billion nucleotide base pairs that make up about 20,000 genes in a single humans genome.7 New AI tools,though,can significantly shorten the time it takes to process,investigate,and provide insights into these large datasets.A particularly promising area where AI is already being ap

27、plied to analyze biodata is precision medicine.A patient can respond very differently to various treatments based on their specific medical history,physical characteristics,and genetic makeup.Precision medicine takes these differences into account,using a patients unique biodata to help decide which

28、 treatments and therapies are likely to work best for that individual.8 For example,knowing the DNA sequence of a patients cancer cells can indicate which genetic mutations led to the disease,and help doctors choose between treatment options that are personalized to the patients specific diagnosis.O

29、verall,medical AI is of great value to precision medicine and public health across a broad range of applications,including:Algorithms to predict how a patient will respond to a specific treatment.AI-guided drug discovery.Integrating electronic health records into genetic databases to connect specifi

30、c DNA mutations to disease outcomes.Using algorithms trained on medical images,such as MRIs or X-rays,to aid doctors in diagnosing patients.Using data from medical records to track the spread of infectious diseases.Center for Security and Emerging Technology|7 Table 1:Types of Biodata Used for Devel

31、oping Next-Generation Technology Applications in the Medical Field Data Example(s)Definition and Function Genomics DNA sequences The human genome is made up of 3 billion base pairs of DNA,some of which encode genes,the instructions for making proteins with specific functions.Mutations,or changes to

32、a gene,can cause a disease or affect how the body works.9 An individuals genetic makeup can impact their response to certain medications,or their likelihood of developing a certain disease.Epigenomics Gene expression,DNA or histone modifications Epigenetics refers to reversible alterations that chan

33、ge how and when certain genes are expressed,or converted to proteins.Typically,epigenetic changes refer to chemically altering DNA or how it is packaged,but do not change a genes DNA sequence.10 Proteomics Protein sequences,functions,and 3D structures Proteins are functional components that perform

34、jobs within cells.Proteins are encoded by genes,which must be built,or expressed.Proteomics studies protein characteristics,presence,location,interactions with other proteins or cellular components,etc.11 Medical Images X-Rays,CT scans,MRIs,ultrasounds Medical images capture visual and electromagnet

35、ic data from patients.12 This information can help to diagnose diseases or medical conditions,or to track how a disease progresses over time.Electronic Medical Records Patient medical history Electronic medical records,or electronic health records,include a patients specific health information like

36、their biometric or demographic information,disease status,and medication history.This data gives insight into the various,complex factors that make up a patients health status.13 Source:CSET analysis.Center for Security and Emerging Technology|8 Medical AIs potential to meet public health needs and

37、overcome cost,labor,and access challenges has many countries striving to harness AIs capabilities in the medical field.14 Many technologically advanced countries around the world,including the United States,the United Kingdom,and European Union,are prioritizing medical AI development through mechani

38、sms including research funding,database maintenance and curation,and collaboration with the private sector.15 For example,U.S.Executive Order 14081 on Advancing Biotechnology and Biomanufacturing Innovation cites the need to“foster a biological data ecosystem that advances biotechnology and biomanuf

39、acturing innovation”to“drive breakthroughs for the United States bioeconomy.”16 While there is a worldwide effort to advance medical AI,Chinas efforts are particularly notable seeing how its government has adopted wide-ranging policies for collection and protection of biodata that could give it an a

40、dvantage in furthering its broader bioeconomy and AI goals.Center for Security and Emerging Technology|9 Policies to Make China a Global Biodata Leader The Chinese government is prioritizing the collection,protection,and promotion of biodata to advance its goals for bioeconomy leadership.While many

41、countries have created strategies to promote AI for public health and precision medicine,medical AI could be especially beneficial for China by solving both urgent and growing public health needs brought on by the countrys large and aging population,as well as its disparate rural healthcare system.1

42、7 China has turned to biodata to develop medical AI for major public health applications like disease surveillance,vaccine development,and food security issues.18 For example,AI tools can reduce the number of physicians needed by helping doctors to diagnose,triage,and predict the best treatment outc

43、omes for patients.19 Tools like digital health services,which integrate medical and clinical information on cloud servers,can help patients in rural areas that might not have access to big medical facilities by reducing in-person hospital visits and streamlining data sharing for physicians to make c

44、linical decisions.20 The Chinese governments central policy planning documents and regulations have consistently prioritized biomedical sciences and AI innovation to pave the way for medical AI.In particular,the Chinese governments commitment to leverage biodata to advance precision medicine,health,

45、and AI is highlighted in the Five-Year Plans and the National Medium-and Long-Term Plan for Science and Technology Development(國家中長期科學和技術發展規劃)as well as in major health-related initiatives like Beijings Healthy China 2030 Policy(“健康中國2030”規劃)(see Appendix A for a table of related Chinese policies,re

46、gulations,and guidelines).21 The Chinese governments inclusion of medical AI development in general country-wide policy documents indicates that precision medicine and emerging medical technologies are a strategic priority.22 That said,there are U.S.initiatives,like the Precision Medicine Initiative

47、(now the National Institutes of Healths“All of Us”program),that similarly aim to support medical AI development.23 The following section examines additional top-down,centralized policies and infrastructure Beijing is putting in place to realize these objectives.In addition to the broad priority-sett

48、ing policies mentioned above,Chinas efforts break down into two categories:1)laws and activities facilitating the collection of data,and 2)laws and activities controlling the access to that data.Policies Facilitating Biodata Accumulation Over the past decade,China has accumulated massive amounts of

49、biodata through a variety of efforts,including instituting requirements to collect and maintain peoples Center for Security and Emerging Technology|10 biodata in big national data centers,digitizing health records,and launching programs to collect biodata from citizens.See Appendix A for a detailed

50、table of many of these policies.Chinese policies that standardize biodata digitization from physical files have facilitated the creation of large datasets that can then be used for research and development.Chinas 14th Five-Year Plan for National Informatization(“十四五”國家信息化規劃)is just one of several po

51、licies that require hospitals to digitize health records and to provide universal digital healthcare,driving further collection of biodata(see Appendix A).24 Additionally,several policies,like Healthy China 2030(“健康中國2030”規劃),instruct standardization of data and connection between data centers,makin

52、g it easier to use across different research applications and specifically directing data collection to promote big data-informed health applications.25 Chinas Biosecurity Law(生物安全法)emphasizes the governments sovereignty over biodata,and the countrys Measures for the Management of Scientific Data(科學

53、數據管理辦法)policy requires that scientific data generated in China considered relevant to national security issues be consolidated in newly developed scientific data centers.26 While some of these measures are not dissimilar from the initiatives countries like the United States and United Kingdom have a

54、dopted to promote biodata collection and consolidation,the Chinese government additionally accumulates biodata through initiatives that may not be considered ethical in other countries and are in fact not standard state uses of biodata,especially in the United States,due to both ethical concerns and

55、 stringent privacy laws.27 According to Human Rights Watch and the Australian Strategic Policy Institute,Chinese government programs have systematically collected millions of peoples biological data,including DNA and blood types,fingerprints,and facial and iris scans,without informed consent or evid

56、ence of criminal activity justifying this collection.28 Programs in Xinjiang and Tibet,for instance,have used alleged anti-terrorism campaigns and medical programs as pretext for collecting biometric data.29 Data from these collection methodologies have been traced to research studies:In 2021,the sc

57、ientific journals Human Genetics and the International Journal of Legal Medicine retracted two scientific publications from Chinese government laboratories which used DNA collected from Uyghur people in Xinjiang and connected genetics to physical appearances,citing“ethical concerns on collection of

58、data.”30 Beijing is focused not only on collecting domestic biodata,but also supports efforts to gain access to biodata from other countries.One example that illustrates Chinas central policy initiatives to accumulate biodata from outside of China is BGI(華大基因),formally the Beijing Genomics Institute

59、,a major genetic sequencing company based in Center for Security and Emerging Technology|11 Shenzhen.In China,BGI runs the government-owned China National GeneBank(CNGB;國家基因庫).31 But subsidiaries of BGI,such as MGI(華大智造),have popped up around the globe.32 In 2012,MGI acquired the California-based co

60、mpany Complete Genomics,which operates DNA sequencing services in direct competition with the U.S.company Illumina.33 Another BGI subsidiary,TECH SOLUTIONS(HONGKONG)CO.,LIMITED,is certified by the U.S.government to provide laboratory testing on patient samples from U.S.hospitals and medical faciliti

61、es.34 The fact that BGI has so many avenues of genetic data collection,across China as well as from other countries through the aforementioned subsidiaries,is concerning for a number of reasons.For one,there are reports that the company holds onto data collected from their services and,on certain oc

62、casions,has also shared biodata collected from non-invasive prenatal tests conducted inside and outside China with Chinas government-funded National GeneBank.35 Moreover,according to media reports,BGI has collaborated frequently with the Peoples Liberation Army,including to develop those prenatal te

63、sts,and the PLA has used at least BGIs domestic Chinese genetic data for military research.36 Policies Controlling Access to Data In tandem with policies designed to promote the large-scale collection of biodata,China is also consolidating its biodata resources by implementing policies and regulatio

64、ns that enhance domestic access and prohibit data from leaving China.Chinese policies requiring the digitization of health records,like the 14th Five-Year Plan for National Informatization(“十四五”國家信息化規劃)mentioned above,aim to make biodata available for researchers in addition to driving more data col

65、lection.37 The CNGB,which contains over 10 million genetic data samples,and the National Genomics Data Center(NGDC 國家基因組科學數據中心)allow researchers in China to access data that can be used to develop medical AI tools.38 This practice is not uncommon and has been implemented in other countries;for examp

66、le,the United States National Institutes of Health(NIH)National Center for Biotechnology Information(NCBI)and the UK Biobank are sources of government-funded and maintained biodata databases for researchers to use.39 However,while researchers in China can access the worlds data through these public

67、databases,the world cannot access Chinas data due to Chinese data privacy laws that prevent foreign researchers from accessing Chinese databases.40 In contrast,U.S.databases like NCBIs GenBank and the Research Collaboratory for Structural Bioinformatics Protein Data Bank(RCSB PDB)do not restrict the

68、 access,use,or distribution of data.41 Center for Security and Emerging Technology|12 Chinas companies also benefit from the countrys biodata holdings and data standardization policies that make it easier for them to access and use that biodata.In contrast,companies in other countries do not have ac

69、cess to biodata to develop technologies in the same manner or at the same scale.In the United States,for example,companies must partner with different hospitals or research institutions to gain access to certain types of biodata for developing medical AI tools,and cannot access this data on their ow

70、n.42 Moreover,U.S.companies face additional challenges because available biodata is often not standardized across different institutions and public health systems,which delays the development of medical AI tools.43 Chinas privacy and security laws directly state that foreign entities cannot use Chin

71、ese genetic data without approval,or export data outside of China,specifically prohibiting institutions and hospitals with any foreign owners or investors from collecting or preserving any human genetic resource from people within China,including data involved in medical device approval processes(se

72、e Appendix A).Chinas Outbound Data Transfer Security Assessment Measures(數據出境安全評估辦法)require that researchers data be assessed by the provincial-level and national cybersecurity and informatization departments before it is published in an academic research paper in a foreign journal.44 Already,these

73、restrictions have forced major scientific publications to grapple with data availability and its relation to transparency and reproducibility.In 2020,for instance,the scientific journal Cell published a study even though the authors,restricted by regulations from the Human Genetic Resource Administr

74、ation of China,refused to deposit the raw genetic data they used for the research.45 By contrast,outside of patient privacy laws like HIPAA,there are no similar restrictions preventing data collected in the United States from being published in foreign journals.46 The above discussion illustrates th

75、at the Chinese government is engaged in a range of efforts to collect massive amounts of biodata,including through unethical means that violate patient privacy as well as by supporting companies that acquire international biodata,as seen with BGI.The Chinese government is also implementing laws that

76、 limit access to Chinese data and prohibit data from leaving the countrys borders.This means that some data collected in China cannot be used to improve technology in other countries or be shared with the international scientific community.The following section elaborates on how China is building an

77、 ecosystem to facilitate the use of this data,including for medical AI applications.Taken together,these measures could give China a competitive advantage in developing new medical AI technologies,which could boost its bioeconomy in comparison to the United States.Center for Security and Emerging Te

78、chnology|13 Policies to Support Research,Development,and Technology Infrastructure To reach its ambitions in biotechnology and AI,Beijing has drafted specific AI-development plans that cross technical areas.These plans urge universities to incorporate AI into biological research and medical healthca

79、re to further basic research in these areas.The 13th Five-Year Plan for the Development of Strategic Emerging Industries(“十三五”國家戰略性新興產業發展規劃)specifically called for co-locating universities and institutes in industrial clusters to“accelerate the pace of innovation and development in the biotech indus

80、try and foster new biotech economic drivers.”47 The AI Innovation Action Plan for Institutions of Higher Education(高等學校人工智能創新行動計劃)states that universities should integrate AI into intelligent medical care and promote biotechnology integration with information technology.48 While these and other gove

81、rnment directives are important to note and track,we must also acknowledge that it is difficult to ascertain their levels of success at this point.49 In addition to supporting academic and government researchers,Beijing also provides financial and other support to private companies as part of its br

82、oader efforts to accelerate medical AI development.For example,in 2010,China Development Bank(國家開發銀行),a state-funded and state-owned bank under the jurisdiction of Chinas State Council,provided$1.5 billion to BGI to extend its research and development platforms.50 More recently,Chinas biomedical ind

83、ustrial clusters,mentioned above,also co-locate private companies alongside the university programs and government research facilities,to support their research.51 Beyond support for research and development,China is also amending some of its medical device policies to accelerate the commercializati

84、on of medical AI technologies.For example,the National Medical Products Administration(NMPA;國家藥品監督管理局)recently reduced barriers to using biodata to develop medical devices by expanding the types of data that can be used to develop technologies.52 This new guidance also reduced certification processi

85、ng time and allows devices to be certified for longer.53 The NMPA regulation opens the door for companies to use a wider variety of health data in their AI technologies,and reduces bureaucratic barriers to apply for approval certification.Chinas Medical AI Research Output Over the past decade,China

86、has nearly matched the United States in the number of medical AI-related English-language research publications per year(Figure 2).In this section we take a closer look at Chinas medical AI research publications over the past decade,compared to the research output from scientists in American institu

87、tions.Center for Security and Emerging Technology|14 Notably,examining publication counts alone does not fully account for a countrys innovative capacity and technological development.The sheer number of papers does not account for the quality or impact of those papers,and so the country with the mo

88、st publications may not be the country making the most influential discoveries.Nevertheless,an assessment of research publication outputs does provide insight into the state of Chinas medical AI research ecosystem and how it has changed and grown over time,partly as a result of the aforementioned bi

89、odata policies the Chinese government is promoting as part of its broader strategy to strengthen the countrys bioeconomy.We drew on CSETs merged corpus of scholarly literature to measure the United States and Chinas predominantly English-language research output in medical AI fields by querying keyw

90、ords that cover both medical and AI topics between 2010 and 2021.54 We found that medical AI research publications from both countries grew steadily over this period,with the United States just barely still leading China by 2021(Figure 2).55 Given the resources the Chinese government is pouring into

91、 this area,Chinas medical AI research output is likely to continue growing.Center for Security and Emerging Technology|15 Figure 2.Trends in Publication Output Containing Medical and AI Keywords for China and the United States,2010-2021 Source:CSETs merged corpus.*For more information on how we gene

92、rated our merged corpus of scholarly literature,see https:/eto.tech/dataset-docs/mac/.Center for Security and Emerging Technology|16 Chinese Companies Developing Biodata-Enabled Medical AI China has been an early adopter of medical AI technology,bolstered by its extensive biodata collection and aggr

93、egation efforts.56 The NMPA has granted several Chinese companies,ranging from medical to technology companies,approvals to use biodata from many sources to develop and deploy medical AI technologies.Below are some examples of Chinese companies that received the NMPAs approval to use patient biodata

94、 to develop medical AI technologies.These companies report a range of successful use cases where theyve deployed medical AI.These claims cannot be taken at face value,since companies have a vested interest in projecting success while not advertising their proprietary capabilities,while Chinas govern

95、ment also benefits from the positive advertising showcasing the countrys biomedical industry.The statements of these companies alone do not prove the success of Chinas medical AI industry.However,the claims mentioned below do provide evidence that companies supported by and working with the Chinese

96、government are using medical AI systems beyond a research setting,which has implications for the bioeconomy worthy of U.S.policymakers attention.Yidu Tech(醫渡科技):According to the companys website,Yidus AI platforms aim to help Chinas hospitals operate more efficiently in medical research,medical mana

97、gement,drug research,and clinical decision support.As of 2021,Yidu claims to have analyzed 1.3 billion medical records across 500 hospitals in China,used AI to help diagnose blood diseases,published over 100 papers using data from their platforms use in hospitals,and provided technology to the Chine

98、se Center for Disease Control and Prevention to assist in pandemic control,assessing transmission risk,and strategizing response.57 Airdoc(鷹瞳科技):The company Airdoc claims to use images of retinas to diagnose various eye diseases,and its AI-based software to diagnose diabetic retinopathy is NMPA-appr

99、oved.The Airdoc algorithms were trained and developed using over 200,000 retinal photographs from 16 clinical settings in China.The companys founder claims this technology has a higher accuracy rate and allows for quicker diagnoses than doctors who rely on alternative tools can provide,and that it i

100、s helping overcome eye doctor shortages in China.58 Infervision(推想科技):Infervision claims to use AI-enabled radiology technology to diagnose cancer through CT scans of lungs,and its InferRead Lung CT.AI technology has received approval for clinical use in the United States and Europe in addition to C

101、hina.59 The companys founder has stated that this Center for Security and Emerging Technology|17 technology will address a major gap in China,as“In China there are just 80,000 radiologists who have to work through 1.4 billion radiology scans every year.”60 According to company statements,Infervision

102、 developed its technology using stored electronic health data to train their algorithms,and incorporated data from the 20 Chinese hospitals to which it was initially deployed.During the COVID-19 pandemic,Infervision translated their lung scan technologies to identify clinical features of COVID-19 an

103、d aid in diagnosing patients.61 Top Chinese technology companies,including those that have not previously been involved in the medical field,are also moving into the medical device space,which further highlights the market potential of successfully translating biodata into cutting-edge medical techn

104、ologies.Some examples of this emerging trend include:Baidu:According to company reporting,Baidus PaddlePaddle(飛獎)AI platform is used in many medical applications from diagnostics to predictions.Doctors use the PaddlePaddle platform to analyze CT scans and detect image patterns to aid in diagnosing p

105、neumonia patients,and most recently,scan the lungs of coronavirus patients for disease.Researchers,including U.S.researchers,used Baidus platform to develop an algorithm that predicts RNA folding structures.62 Baidu also states that their platform was used during the COVID-19 pandemic to power predi

106、ctions of coronavirus RNA structure for future pharmaceutical applications.63 Alibaba:In addition to working with BGI to provide computing power and storage for massive amounts of genetic data,Alibaba also developed their ET Medical Brain(ET醫療大腦)AI platform.The company claims the platform can be use

107、d to integrate hospital logistics and diagnostic assistance for better patient outcomes by incorporating data from patient electronic health records,and patient images.64 Tencent:After being tapped by the Ministry of Science and Technology(MOST)in 2017 as a“National AI Team”member,earning government

108、 support and access to regional projects and data resources to construct AI platforms for medical imaging,Tencent announced that they launched the Jarvis Lab(騰訊天衍研究中心)in 2018 to focus on developing Medical Artificial Intelligence.According to Tencent,the AI products from this laboratory are used to

109、integrate electronic medical records and disease risk prediction,and automatically analyze medical images such as CT and PET scans to aid doctors in diagnostics.Tencent has also reported a partnership with pharmaceutical giant Novartis on an AI Center for Security and Emerging Technology|18 Nurse pl

110、atform for digital health access in China,and claims to be working to make their AI-powered digital health service WeDoctor(微醫)public.65 The companies involved in medical AI development and deployment have also faced some challenges.For instance,Baidu,Alibaba,and Tencent(BAT)have reportedly had trou

111、ble gaining regulatory approval from the NMPA,unlike many smaller players in the medical AI space who have had more success due to their increased strategic focus on medical AI compared to BATs more disparate priorities.66 However,the efforts of a variety of Chinese companies to enter this industry

112、reflect Chinas push towards medical AI.Center for Security and Emerging Technology|19 Conclusions and Implications of Chinas Biodata Leadership China is making rapid progress in developing and implementing medical AI technologies,in part because of national strategies that facilitate the collection

113、and integration of biodata.Beijings dozens of dedicated policies to collect and use health data demonstrate its prioritization of this issue,and are likely to lead to continued growth in medical AI research.As China develops these tools and brings them to international markets,U.S.policymakers will

114、have to contend with various challenges.In particular,it is important to consider how Chinas accumulation and use of biodata impacts the United States competitiveness in the global bioeconomy.Chinese researchers have access to far more biodata than U.S.researchers do,including large public datasets

115、from the United States,Europe,and others,plus all of the genetic data accumulated within China that U.S.researchers do not have access to.While nothing is for certain,this comparative advantage could help China become a global powerhouse in medical AI and capture the economic value that comes with i

116、t.Furthermore,whoever dominates the space of next-generation medical technologies will have an outsized impact on research standards and norms,and on how future developments and applications are prioritized.If China becomes the world leader in this arena,future developments may not align with U.S.pr

117、iorities due to differences in research transparency,privacy,and medical device approval mechanisms.Major Chinese technology champions are already permeating the global health space;for example,Infervisions“InferRead Lung CT.AI”tool is approved as a medical device in the United States by the U.S.Foo

118、d and Drug Administration.67 When medical AI technologies from Chinese companies are approved and adopted in the United States,health data from U.S.patients could be collected and fed back into Chinas databases.While this approach to biodata collection and accumulation could lead to more efficient a

119、nd beneficial medical AI,it would also widen the data resource gap between China and the rest of the world.In summary,the Chinese government has created a unique ecosystem that treats biodata as a resource,and accumulates and uses it to advance medical AI development.This centralized strategy may gi

120、ve China an edge against the United States,and its significant economic,societal,and technological implications should be considered in future policies that aim to bolster the U.S.bioeconomy.Center for Security and Emerging Technology|20 Appendix A:Select List of Chinas Notable Governmental Policies

121、 on Medical AI Policy Year Implications Guidelines for the Registration and Examination of AI Medical Devices (人工智能醫療器械注冊審查指導原則)68 2022 Introduces a more detailed definition of allowed data sources for AI devices,expanding the scope from“data from medical devices”to encompass medical and non-medical

122、 information,including medical records,exam results,and patient complaints,etc.Expedites process and certification length.Outbound Data Transfer Security Assessment Measures(數據出境安全評估辦法)69 2022 Requires oversight and restricts data of any kind leaving China by conducting security assessments.Requires

123、 oversight of any researchers data before using it in an academic research paper in a foreign journal.14th Five-Year Plan for National Economic and Social Development and Long-Range Objectives for 2035(中華人民共和國國民經濟和社會發展第十四個五年規劃和2035年遠景目標)70 2021 Calls to:Apply genetic research for technological innov

124、ations.Establish national laboratories with a focus on AI,biotechnology,and pharmaceuticals.Reorganize the state key laboratories to form a laboratory system.Name biotechnology a strategic emerging industry.Accelerate deployment of cutting-edge technologies.Personal Information Protection Law(中華人民共和

125、國個人信息保護法)71 2021 Protects individuals privacy by requiring consent to collect or process personal data,including medical and biometric information.Prevents personal data from transferring outside of China.Center for Security and Emerging Technology|21 14th Five-Year Plan for National Informatization

126、(“十四五”國家信息化規劃)72 2021 Requires hospitals to digitize health records.Provides universal digital healthcare.Establish standardization of data.Guiding Opinions on Expanding Investment in Strategic Emerging Industries and Cultivating Strengthened New Growth Points and Growth Poles(關于擴大戰略性新興產業投資 培育壯大新增長點

127、增長極的指導意見)73 2020 Calls to:Accelerate pace of innovation in the biotechnology industry.Establish biotechnology and pharmaceutical innovation centers.Create markets for pharmaceutical drugs and medical equipment.Biosecurity Law(中華人民共和國生物安全法)74 2020 Strengthens the management and oversight of the colle

128、ction,storage,use,and external provision of Human Genetic Resources.Allows the government to“enjoy sovereignty”over Human Genetic Resources and biological resources.Regulations on the Administration of Human Genetic Resources(中華人民共和國人類遺傳資源管理條例)75 2019 Organizations must apply for a certification to

129、use Human Genetic Resources within China.Prohibits collection or preservation of Human Genetic Resources by foreign organizations and individuals.Foreign organizations must have a domestic sponsor to apply or use any Human Genetic Resources within China.Measures for the Management of Scientific Data

130、(科學數據管理辦法)76 2018 Requires scientific data generated in China that is considered relevant to national security issues be consolidated in newly-developed scientific data centers.Center for Security and Emerging Technology|22 Opinions on Promoting the Development of“Internet+Healthcare”(促進“互聯網+醫療健康”發展

131、的意見)77 2018 Enhances public health services by integrating technology into hospitals and public health.Part of the“Internet+”initiative.AI Innovation Action Plan for Institutions of Higher Education(高等學校人工智能創新行動計劃)78 2018 Calls universities to:Promote the deep integration of information technologies

132、 with modern biotechnology.Apply AI to intelligent medical care.New Generation Artificial Intelligence Development Plan(新一代人工智能發展規劃)79 2017 Directs organizations to develop new models and methods of AI in fields including healthcare,precision medicine,and genetics.Chinese Population Precision Medici

133、ne Research Program(中國人群精準醫學研究計劃)80 2016 Funds a$9.2 billion,15-year program to“build up the countrys credentials in precision medicine.”Establishes precision medicine centers between hospitals,universities,and sequencing companies.Plans to sequence over 100 million human genomes to beat the same go

134、al as the United States Precision Medicine Initiative.Establish a database of genomic data from Chinese and international populations.Healthy China 2030 Plan (“健康中國2030”規 劃)81 2016 Calls to:Promote the construction of a healthy China through reform,innovation,and scientific development.Collect,integ

135、rate,and share data to promote big data-informed health applications.Center for Security and Emerging Technology|23 National 13th 5-Year Plan for S&T Innovation(“十三五”國家科技創新規劃)82 2016-2020 Calls to:Develop cutting-edge and general-purpose biotechnology,including gene editing and synthetic biology.Sei

136、ze a commanding position in international biotechnology competition.National 13th Five-Year Plan for the Development of Strategic Emerging Industries(“十三五”國家戰略性新興產業發展規劃)83 2016-2020 Calls to:Accelerate development of genomics and associated technologies for use in precision medicine,biosynthesis,and

137、 agriculture.Promote disruptive technological innovation in agricultural biology.Promote the clinical application and industrialization of artificial biology and artificial biological devices.Made in China 2025(中國制造2025)84 2015 Calls to:Integrate new generation information technology and manufacturi

138、ng,including in science and technology innovations like bioengineering.The National Medium-and Long-Term Program for Science and Development (國家中長期科學和技術發展規劃)85 2006-2020 Prioritizes development of biotechnology and life sciences.Establishes mega-projects to meet development goals of China,including

139、in public health and genetically-modified agriculture.Calls for research and development of genome sequencing and genetic structure analysis.Calls for breakthroughs in the fields of functional genome,proteomics,stem cells,etc.Source:CSET analysis.Current as of September 2023.Center for Security and

140、Emerging Technology|24 Authors Caroline Schuerger is a former research fellow at the Center for Security and Emerging Technology(CSET),where Vikram Venkatram is a research analyst,and Katherine Quinn is a data scientist.Acknowledgments For feedback and assistance,we would like to thank Anna Puglisi,

141、Steph Batalis,Daniel Chou,Emily Weinstein,Catherine Aiken,Owen Daniels,Ben Murphy,Shelton Fitch,Jason Ly,Margarita Konaev,and Igor Mikolic-Torreira.The authors are especially grateful to Jeff Ding and Mark Kazmierczak for their comprehensive reviews.2024 by the Center for Security and Emerging Techn

142、ology.This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License.To view a copy of this license,visit https:/creativecommons.org/licenses/by-nc/4.0/.Document Identifier:doi:10.51593/20220018 Center for Security and Emerging Technology|25 Endnotes 1 Sean Mann,

143、Carl T.Berdahl,Lawrence Baker,Federico Girosi,and Rand Health,“Applications Used in the Clinical and Public Health Response to COVID-19,”Patient-Centered Outcomes Research Institute,December 2022,https:/www.pcori.org/sites/default/files/PCORI-Emerging-Technologies-Therapeutics-Reports-AI-In-Response

144、-COVID-19-Volume-1.pdf.2 Marcy E Gallo,“The Bioeconomy:A Primer,”Congressional Research Service,September 22,2022,https:/crsreports.congress.gov/product/pdf/R/R46881/4;Market Research Future“Artificial Intelligence in Healthcare Market Size Worth USD 12.22 Billion by 2030 at 51.9%CAGR Report by Mark

145、et Research Future(MRFR),”GlobalNewswire,January 24,2023,https:/ Chui,Matthias Evers,James Manyika,Alice Zheng,and Travers Nisbet,“The Bio Revolution:Innovations transforming economies,societies,and our lives,”McKinsey&Company,May 13,2020,https:/ Ben Buchanan,“The AI Triad and What It Means for Nati

146、onal Security Strategy,”Center for Security and Emerging Technology,August 2020,https:/cset.georgetown.edu/publication/the-ai-triad-and-what-it-means-for-national-security-strategy/.4 Bethany Allen-Ebrahimian,“China Makes Genetic Data a National Resource,”Axios,March 29,2022,https:/ Yeung,“Chinas Si

147、tting on a Goldmine of Genetic Data and It Doesnt Want to Share,”CNN,August 12,2023,https:/ CSET Translation of“National 13th Five-Year Plan for the Development of Strategic Emerging Industries”(“十三五”國家戰略性新興產業發展規劃),State Council(國務院),November 29,2016,https:/cset.georgetown.edu/publication/national-1

148、3th-five-year-plan-for-the-development-of-strategic-emerging-industries/;Infervision,“How AI And Deep Learning Are Now Used To Diagnose Cancer-Infervision,”accessed September 5,2023,https:/ Wilson Wen Bin Goh,and Limsoon Wong,“The Birth of Bio-Data Science:Trends,Expectations,and Applications,”Genom

149、ics,Proteomics&Bioinformatics,February 2020,https:/doi.org/10.1016/j.gpb.2020.01.002.7 Genome.gov,“Introduction to Genomics,”accessed September 5,2023,https:/www.genome.gov/About-Genomics/Introduction-to-Genomics.Center for Security and Emerging Technology|26 8“The Promise of Precision Medicine,”Nat

150、ional Institutes of Health,February 5,2020,https:/www.nih.gov/about-nih/what-we-do/nih-turning-discovery-into-health/promise-precision-medicine.9“Introduction to Genomics:Whats a Genome?”National Institutes of Health National Human Genome Research Institute,accessed September 5,2023,https:/www.genom

151、e.gov/About-Genomics/Introduction-to-Genomics;“Artificial Intelligence,Machine Learning and Genomics,”National Institutes of Health National Human Genome Research Institute,accessed September 5,2023,https:/www.genome.gov/about-genomics/educational-resources/fact-sheets/artificial-intelligence-machin

152、e-learning-and-genomics;iga Avsec,“Predicting Gene Expression with AI,”Google DeepMind,October 4,2021,https:/ Mohamad M.Kronfol,Mikhail G.Dozmorov,Rong Huang,Patricia W.Slattum,and Joseph L.McClay,“The Role of Epigenomics in Personalized Medicine,”Expert Review of Precision Medicine and Drug Develop

153、ment,January 17,2017,https:/doi.org/10.1080/23808993.2017.1284557;Lawrence B.Holder,M.Muksitul Haque,and Michael K.Skinner,“Machine Learning for Epigenetics and Future Medical Applications,”Epigenetics:Official Journal of the DNA Methylation Society June 21,2017,https:/doi.org/10.1080/15592294.2017.

154、1329068.11“What Is Proteomics?”EMBL-EBI,accessed September 5,2023,https:/www.ebi.ac.uk/training/online/courses/proteomics-an-introduction/what-is-proteomics/.12“Medical Imaging,”U.S.Food and Drug Administration,2018,https:/www.fda.gov/radiation-emitting-products/radiation-emitting-products-and-proce

155、dures/medical-imaging;“More about MIDRC,”Medical Imaging and Data Resource Center,accessed September 5,2023,https:/www.midrc.org/more-about-midrc.13“Electronic Health Records,”Centers for Medicare and Medicaid Services,accessed September 5,2023,https:/www.cms.gov/medicare/e-health/ehealthrecords.14

156、Sarah De Nigris,Massimo Craglia,Daniel Nepelski,Jiri Hradec,Emilio Gomez-Gonzales,Emilia Gomez Gutierrez,Miguel Vazquez-Prada Baillet,Riccardo Righi,Guiditta De Prato,Montserrat Lopez Cobo,Sofia Samoili,and Melisande Cardona,“AI Watch:AI Uptake in Health and Healthcare,2020,”Publications Office of t

157、he European Union,2020,ISBN 978-92-76-26936-6,doi:10.2760/948860,JRC122675;Angela Spatharou,Solveigh Hieronimus,and Jonathan Jenkins,“Transforming Healthcare with AI:The Impact on the Workforce and Organizations,”McKinsey&Company,March 10,2020,https:/ UK Department for Science,Innovation,and Technol

158、ogy,“New 100 million fund to capitalise on AIs game-changing potential in life sciences and healthcare,”October 29,2023,Center for Security and Emerging Technology|27 https:/www.gov.uk/government/news/new-100-million-fund-to-capitalise-on-ais-game-changing-potential-in-life-sciences-and-healthcare;L

159、ael Brainard,Neera Tanden,and Arati Prabhakar,“Delivering on the Promise of AI to Improve Health Outcomes,”White House,December 14,2023,https:/www.whitehouse.gov/briefing-room/blog/2023/12/14/delivering-on-the-promise-of-ai-to-improve-health-outcomes/;“EU invests 220 million to test AI solutions for

160、 healthcare,food,industry,and everyday life,”European Commission,June 29,2023,https:/digital-strategy.ec.europa.eu/en/news/eu-invests-eu220-million-test-ai-solutions-healthcare-food-industry-and-everyday-life.16 The White House,“Executive Order on Advancing Biotechnology and Biomanufacturing Innovat

161、ion for a Sustainable,Safe,and Secure American Bioeconomy,”September 12,2022,https:/www.whitehouse.gov/briefing-room/presidential-actions/2022/09/12/executive-order-on-advancing-biotechnology-and-biomanufacturing-innovation-for-a-sustainable-safe-and-secure-american-bioeconomy/.17 World Bank Open Da

162、ta,“Physicians(per 1,000 People.),”accessed September 5,2023,https:/data.worldbank.org/indicator/SH.MED.PHYS.ZS?locations=CN-US-AU-BR-TR;Sarah OMeara,“Chinas Data-Driven Dream to Overhaul Health Care,”Nature 598(7879):S13,2021,https:/ World Food Programme,“China,”accessed September 5,2023,https:/www

163、.wfp.org/countries/china;Ethan Wang and Bernard Orr,“China to Promote Vaccine Development,New Medicines,”Reuters,March 5,2023,https:/ Patton,“Exclusive:China Rolls out GMO Corn Planting,Starts Small,”Reuters,February 17,2023,https:/ Wang,Gong-Huan Yang,Jia-Qi Ma,Li-Ping Wang,and Xiao-Peng Qi.,“China

164、 Information System for Disease Control and Prevention(CISDCP),”National Bureau of Asian Research,accessed September 5,2023,https:/www.nbr.org/wp-content/uploads/pdfs/programs/CISDCP.pdf.19“Yidu Cloud,”accessed September 5,2023,https:/ Lulu Yilun Chen,“Tencent-Backed WeDoctor Plans IPO Filing by End

165、 of April”Bloomberg,January 8,2023,https:/ Through Healthy China 2030,the government has reportedly invested over$9 billion USD for the Chinese Population Precision Medicine Research Program(中國人群精準醫學研究計劃)to utilize the wealth of accumulated health biodata,and to“realize data collection,integration,s

166、haring,and business of public health.”International Telecommunication Union(ITU)translation of“The National Medium-and Center for Security and Emerging Technology|28 Long-Term Program for Science and Technology Development(2006-2020):An Outline”(國家中長期科學和技術發展規劃綱要（20062020年）),State Council,2006,https:

167、/www.itu.int/en/ITU-D/Cybersecurity/Documents/National_Strategies_Repository/China_2006.pdf;“The Chinese Communist Party Central Committee and the State Council Issue Outline of the Healthy China 2030 Plan”(中共中央 國務院印發“健康中國2030”規劃綱要),Xinhua News Agency,October 25,2016,https:/perma.cc/8576-U8HX.22 Dav

168、id Cyranoski,“China Embraces Precision Medicine on a Massive Scale.”Nature 529(7584):910,January 6,2016,https:/doi.org/10.1038/529009a.23“All of Us Research Program Overview,”National Institute of Health All of Us Research Program,June 22,2020,https:/allofus.nih.gov/about/program-overview.24“Transla

169、tion:14th Five-Year Plan for National InformatizationDec.2021,”DigiChina,https:/digichina.stanford.edu/work/translation-14th-five-year-plan-for-national-informatization-dec-2021/.25“The Chinese Communist Party Central Committee and the State Council Issue Outline of the Healthy China 2030 Plan”(中共中央

170、 國務院印發“健康中國2030”規劃綱要),Xinhua News Agency,October 25,2016,https:/perma.cc/8576-U8HX.26 China Law Translate translation of“Biosecurity Law of the P.R.C.”(中華人民共和國生物安全法),National Peoples Congress Standing Committee(國家人民代表大會常務委員會),October 17,2020,https:/perma.cc/VM3X-DW9X;CSET translation of“Measures for

171、 the Management of Scientific Data”(科學數據管理辦法),General Office of the State Council(國務院辦公廳),April 2,2018,https:/cset.georgetown.edu/publication/china-scientific-data-management-measures/.27 “Privacy in Genomics,”National Institutes of Health National Human Genome Research Institute,accessed September

172、5,2023,https:/www.genome.gov/about-genomics/policy-issues/Privacy.28 Emile Dirks and James Leibold,“Genomic Surveillance,”Australian Strategic Policy,June 17,2020,https:/www.aspi.org.au/report/genomic-surveillance;Sophie Richardson,“China:Minority Region Collects DNA from Millions,”Human Rights Watc

173、h,December 13,2017,https:/www.hrw.org/news/2017/12/13/china-minority-region-collects-dna-millions.29 United States Department of the Treasury,Office of Foreign Assets,“Treasury Sanctions Chinese Entity and Officials Pursuant to Global Magnitsky Human Rights Executive Order,”July 30,2020,https:/home.

174、treasury.gov/news/press-releases/sm1073;Sophie Richardson,“China:Minority Region Collects DNA from Millions,”Human Rights Watch,December 13,2017,https:/www.hrw.org/news/2017/12/13/china-minority-region-collects-dna-millions;Emile Dirks and James Leibold,“Genomic Surveillance,”Australian Strategic Po

175、licy,June 17,2020,Center for Security and Emerging Technology|29 https:/www.aspi.org.au/report/genomic-surveillance;Emile Dirks,“Mass DNA Collection in the Tibet Autonomous Region from 20162022,”The Citizen Lab,September 13,2022,https:/citizenlab.ca/2022/09/mass-dna-collection-in-the-tibet-autonomou

176、s-region/.30 Yi Li,Wenting Zhao,Dan Li,Xianming Tao,Ziyi Xiong,Jing Liu,Wei Zhang,et al.,“RETRACTED ARTICLE:EDAR,LYPLAL1,PRDM16,PAX3,DKK1,TNFSF12,CACNA2D3,and SUPT3H Gene Variants Influence Facial Morphology in a Eurasian Population,”Human Genetics 138(6):68189 April 25,2019,https:/doi.org/10.1007/s

177、00439-019-02023-7;Xiaoxi Jing,Yanan Sun,Wenting Zhao,Xingjian Gao,Mi Ma,Fan Liu,and Caixia Li,“RETRACTED ARTICLE:Predicting Adult Height from DNA Variants in a European-Asian Admixed Population,”International Journal of Legal Medicine 133(6):166779,April 12,2019,https:/doi.org/10.1007/s00414-019-020

178、39-8.31“China National GeneBank(CNGB)-About”BGI Group Official Website,accessed September 5,2023,https:/ MGI Tech,“MGI-Genetic Sequencing and Multi-Omics Solutions for Life Science and Healthcare Research,”accessed September 5,2023,https:/en.mgi- Genomics,BGI Agree to$117.6M Merger,”GenomeWeb,Septem

179、ber 17,2012,https:/ Current CLIA Identification number is 99D2135851,and expires on 09/07/2025;Reuters.“BRIEF-BGI Genomics Unit Receives CLIA Certificate,”September 11,2017,https:/ for Medicare and Medicaid Services,“Clinical Laboratory Improvement Amendments(CLIA),”accessed September 5,2023,https:/

180、www.cms.gov/regulations-and-guidance/legislation/clia.35 Kirsty Needham,“Exclusive:China Gene Firm Providing Worldwide COVID Tests Worked with Chinese Military,”Reuters,January 30,2021,https:/ Needham and Clare Baldwin,“Chinas Gene Giant Harvests Data from Millions of Pregnant Women,”Reuters,July 7,

181、2021,https:/ Kirsty Needham,“Exclusive:China Gene Firm Providing Worldwide COVID Tests Worked with Chinese Military.”;Kirsty Needham and Clare Baldwin,“Chinas Gene Giant Harvests Data from Millions of Pregnant Women.”37“Translation:14th Five-Year Plan for National InformatizationDec.2021,”DigiChina,

182、https:/digichina.stanford.edu/work/translation-14th-five-year-plan-for-national-informatization-dec-2021/.Center for Security and Emerging Technology|30 38 Chinese Academy of Sciences,“Big Gene Bank to Anchor Precision Medicine,”Chinese Academy of Sciences,December 9,2015,https:/ National Genebank,N

183、ational Genomics Data Center,“About-National Genomics Data Center(CNCB-NGDC),”accessed September 5,2023,https:/ Overview,”National Institutes of Health,National Library of Medicine,National Center for Biotechnology Information,accessed September 5,2023,https:/www.ncbi.nlm.nih.gov/genbank/;UK Biobank

184、,“About UK Biobank,”accessed September 5,2023,https:/www.ukbiobank.ac.uk/.40 Jessie Yeung,“Chinas Sitting on a Goldmine of Genetic Data-and It Doesnt Want to Share,”CNN,August 12,2023,https:/ Overview,”National Institutes of Health,National Library of Medicine,National Center for Biotechnology Infor

185、mation,accessed January 19,2024,https:/www.ncbi.nlm.nih.gov/genbank/;RCSB Protein Data Bank,“About RCSB PDB:Enabling Breakthroughs in Scientific and Biomedical Research and Education,”accessed January 19,2024,https:/www.rcsb.org/pages/about-us/index.42 Center for Open Data Enterprise,“Sharing and Ut

186、ilizing Health Data for AI Applications:Roundtable Report,”2019,https:/www.hhs.gov/sites/default/files/sharing-and-utilizing-health-data-for-ai-applications.pdf.43 Eric Wu,Kevin Wu,Roxana Daneshjou,David Ouyang,Daniel E.Ho,and James Zou,“How Medical AI Devices Are Evaluated:Limitations and Recommend

187、ations from an Analysis of FDA Approvals,”Nature Medicine 27(4):58284,2017,April 5,2021,https:/doi.org/10.1038/s41591-021-01312-x;Shana Lynch,“The Geographic Bias in Medical AI Tools,”Stanford University Human-Centered Artificial Intelligence,September 21,2020,https:/hai.stanford.edu/news/geographic

188、-bias-medical-ai-tools;Center for Open Data Enterprise,“Sharing and Utilizing Health Data for AI Applications:Roundtable Report.”44“Translation:Outbound Data Transfer Security Assessment Measures Effective Sept.1,2022,”DigiChina,https:/digichina.stanford.edu/work/translation-outbound-data-transfer-s

189、ecurity-assessment-measures-effective-sept-1-2022/.45 Jun-Yu Xu,Chunchao Zhang,Xiang Wang,Linhui Zhai,Yiming Ma,Yousheng Mao,Kun Qian,et al.,“Integrative Proteomic Characterization of Human Lung Adenocarcinoma,”Cell 182(1):245-261.e17,July 19,2020,https:/doi.org/10.1016/j.cell.2020.05.043.Center for

190、 Security and Emerging Technology|31 46 Office for Civil Rights(OCR),“Your Rights under HIPAA.”U.S.Department of Health and Human Services,May 7,2008,https:/www.hhs.gov/hipaa/for-individuals/guidance-materials-for-consumers/index.html.47 CSET translation of“National 13th Five-Year Plan for the Devel

191、opment of Strategic Emerging Industries”(“十三五”國家戰略性新興產業發展規劃),State Council(國務院),November 29,2016,https:/cset.georgetown.edu/publication/national-13th-five-year-plan-for-the-development-of-strategic-emerging-industries/.48 CSET translation of“Notice of the Ministry of Education on the Publication of

192、the AI Innovation Action Plan for Institutions of Higher Education”(教育部關于印發高等學校人工智能創新行動計劃的通知),April 2,2018,https:/cset.georgetown.edu/research/ai-innovation-action-plan-for-institutions-of-higher-education.49 Anna Puglisi and Daniel Chou,“Chinas Industrial Clusters:Building AI-Driven Bio-Discovery C

193、apacity,”Center for Security and Emerging Technology,June 2022,https:/cset.georgetown.edu/publication/chinas-industrial-clusters/;Emily S.Weinstein,Channing Lee,Ryan Fedasiuk,and Anna Puglisi,“Chinas State Key Laboratory System:A View into Chinas Innovation System,”Center for Security and Emerging T

194、echnology,June 2022,https:/cset.georgetown.edu/publication/chinas-state-key-laboratory-system/.50 Julia Karow,“BGI to Receive$1.5B in collaborative Funds over 10 Years from China Development Bank,”GenomeWeb,January 12,2010,https:/ Anna Puglisi and Daniel Chou,“Chinas Industrial Clusters:Building AI-

195、Driven Bio-Discovery Capacity.”52The National Medical Products Administration(NMPA;國家藥品監督管理局)is the agency in China that is responsible for approving and certifying medical devices and pharmaceutical drugs.53 National Medical Products Association,“Key Points of Deep Learning Decision-Assisting Medic

196、al Device Software Review”Archived September 11,2022,https:/perma.cc/3VVC-7J3B.54 CSET merged corpus of scholarly literature including Web of Science,OpenAlex,Semantic Scholar,The Lens,arXiv,and Papers With Code.Certain data included herein are derived from Clarivate Web of Science.Copyright Clariva

197、te 2023.All rights reserved.The majority of publications identified from our keyword search were English-language,though we did include Chinese-language keywords and the search did find some relevant papers in Chinese and other languages.Center for Security and Emerging Technology|32 55 A publicatio

198、n is counted for a country if at least one of its authors is affiliated with a research institution in that country.For example,a paper with one author affiliated with a Chinese institution and one author affiliated with a U.S.institution is added as one to each countrys paper count.For more discuss

199、ion on this counting method,see Jacob Feldgoise,Catherine Aiken,Emily S.Weinstein,and Zachary Arnold,“Studying Tech Competition through Research Output:Some CSET Best Practices,”Center for Security and Emerging Technology,April 2023,https:/cset.georgetown.edu/article/studying-tech-competition-throug

200、h-research-output-some-cset-best-practices/.56 Eleanor Olcott,“China Sets the Pace in Adoption of AI in Healthcare Technology,”Financial Times,January 31,2022,https:/ Nature Advertisement,“Spearheading a New Age of Medical Intelligence,”Nature,accessed September 5,2023,https:/ Cloud,“Medical Big Dat

201、a Intelligent Solutions,”accessed September 5,2023,https:/ Cloud“Yidu Tech Announces Annual Results for Fiscal Year 2022,”accessed September 5,2023,https:/ Cloud,“Yidu Tech Leads Chinas Healthcare Intelligence Market,”accessed September 5,2023,https:/ Microsoft,“AI and Preventative Healthcare:Diagno

202、sis in the Blink of an Eye,”Microsoft Stories Asia,September 17,2018,https:/ Lin,Jianhao Xiong,Congxin Liu,Lanqin Zhao,Zhongwen Li,Shanshan Yu,Xiaohang Wu,et al.,“Application of Comprehensive Artificial Intelligence Retinal Expert(CARE)System:A National Real-World Evidence Study.”The Lancet.Digital

203、Health 3(8):e48695,August 2021,https:/doi.org/10.1016/s2589-7500(21)00086-8;National Medical Products Administration,accessed September 2023,https:/ InferVision,“Infervision Receives FDA Clearance for the InferRead Lung CT.AI Product,”accessed January 24,2024,https:/ Infervision,“How AI And Deep Lea

204、rning Are Now Used To Diagnose Cancer-Infervision,”accessed September 5,2023,https:/ Infervision,“How AI And Deep Learning Are Now Used To Diagnose Cancer-Infervision”;Emily S.Weinstein,“Chinas Use of AI in its COVID-19 Response,”Center for Security and Emerging Technology,August 2020,https:/cset.ge

205、orgetown.edu/publication/chinas-use-of-ai-in-its-covid-19-response/.Center for Security and Emerging Technology|33 62 Baidu,“How Baidu Is Bringing AI to the Fight against Coronavirus,”MIT Technology Review,March 11,2020,https:/ Baidu,“Baidus Deep-Learning Platform Fuels the Rise of Industrial AI,”MI

206、T Technology Review,June 22,2020,https:/ Baidu Is Bringing AI to the Fight against Coronavirus”;Liang Huang,He Zhang,Dezhong Deng,Kai Zhao,Kaibo Liu,David A.Hendrix,and David H.Mathews,“LinearFold:Linear-Time Approximate RNA Folding by 5-to-3 Dynamic Programming and Beam Search,”Bioinformatics(Oxfor

207、d,England)35(14):i295304 July 2019,https:/doi.org/10.1093/bioinformatics/btz375.64 NASDAQ,“How Alibaba Is Using Artificial Intelligence In Healthcare,”July 13,2017,https:/ Cloud“Medical Brain:Data Intelligence for Healthcare&Lives Saving-Alibaba Cloud,”accessed September 5,2023,https:/ Tencent,“Jarv

208、is Lab,”accessed September 5,2023,https:/ Yilun Chen,“Tencent-Backed WeDoctor Plans IPO Filing by End of April”Bloomberg,January 8,2023,https:/ Nurse Set to Transform Standards of Care for Heart Failure Patients in China,”April 30,2020,https:/ Announces AIMIS Medical Image Cloud and AIMIS Open Lab H

209、elp Medical Data Management and Accelerate Incubation of Medical AI Application,”PR Newswire,October 20,2020,https:/ Larsen,“Drafting Chinas National AI Team for Governance,”DigiChina,November 18,2019,https:/digichina.stanford.edu/work/drafting-chinas-national-ai-team-for-governance/.66 Yuchen Li,“T

210、he Misfires:How BAT All Stumbled in Medical AI,”Translated by Wenmiao Liu,August 6 2021,Leiphone,https:/ U.S.Food and Drug Administration,“Artificial Intelligence and Machine Learning(AI/ML)-Enabled Medical Devices,”October 5,2022,https:/www.fda.gov/medical-devices/software-medical-device-samd/artif

211、icial-intelligence-and-machine-learning-aiml-enabled-medical-devices.Center for Security and Emerging Technology|34 68 China National Medical Products Administration,“Guidelines for the Registration and Examination of AI Medical Devices,”https:/perma.cc/U8T5-9N4U.69“Translation:Outbound Data Transfe

212、r Security Assessment Measures Effective Sept.1,2022,”DigiChina,https:/digichina.stanford.edu/work/translation-outbound-data-transfer-security-assessment-measures-effective-sept-1-2022/.70 CSET translation of“Outline of the Peoples Republic of China 14th Five-Year Plan for National Economic and Soci

213、al Development and Long-Range Objectives for 2035”(中華人民共和國國民經濟和社會發展第十四個五年規劃和2035年遠景目標綱),Xinhua News Agency(新華社),March 12,2021,https:/cset.georgetown.edu/publication/china-14th-five-year-plan.71“Translation:Personal Information Protection Law of the Peoples Republic of ChinaEffective Nov.1,2021,”Digi

214、China blog,https:/digichina.stanford.edu/work/translation-personal-information-protection-law-of-the-peoples-republic-of-china-effective-nov-1-2021.72“Translation:14th Five-Year Plan for National InformatizationDec.2021,”DigiChina,https:/digichina.stanford.edu/work/translation-14th-five-year-plan-fo

215、r-national-informatization-dec-2021/.73 CSET-DigiChina translation of“Guiding Opinions on Expanding Investment in Strategic Emerging Industries and Cultivating Strengthened New Growth Points and Growth Poles”(關于擴大戰略性新興產業投資 培育壯大新增長點增長極的指導意見),National Development and Reform Commission(國家發展和改革委員會),Sept

216、ember 8,2020,https:/cset.georgetown.edu/publication/new-chinese-ambitions-for-strategic-emerging-industries-translated.74 China Law Translate translation of“Biosecurity Law of the P.R.C.”(中華人民共和國生物安全法),National Peoples Congress Standing Committee(國家人民代表大會常務委員會),October 17,2020,https:/perma.cc/VM3X-D

217、W9X.75 Ministry of Science and Technology(MOST;科學技術部),“Regulations on the Administration of Human Genetic Resources of the Peoples Republic of China”(中華人民共和國人類遺傳資源管理條例),May 28,2019,https:/perma.cc/WZ76-K3R6.76 CSET translation of“Measures for the Management of Scientific Data”(科學數據管理辦法),General Offi

218、ce of the State Council(國務院辦公廳),April 2,2018,https:/cset.georgetown.edu/publication/china-scientific-data-management-measures/.Center for Security and Emerging Technology|35 77“Opinions of the General Office of the State Council on Promoting the Development of Internet+Healthcare”(國務院辦公廳關于促進“互聯網+醫療健

219、康”發展的意見),April 25,2018,https:/perma.cc/47TT-YGAG.78 CSET translation of“Notice of the Ministry of Education on the Publication of the AI Innovation Action Plan for Institutions of Higher Education”(教育部關于印發高等學校人工智能創新行動計劃的通知),April 2,2018,https:/cset.georgetown.edu/research/ai-innovation-action-plan-f

220、or-institutions-of-higher-education.79“Full Translation:Chinas New Generation Artificial Intelligence Development Plan(2017),”DigiChina blog,https:/digichina.stanford.edu/work/full-translation-chinas-new-generation-artificial-intelligence-development-plan-2017/.80 Brian Wang,“Chinas$9.2 Billion Prec

221、ision Medicine Initiative Could See about 100 Million Whole Human Genomes Sequenced by 2030 and More If Sequencing Costs Drop,”N,June 7,2016;and David Cyranoski,“China Embraces Precision Medicine on a Massive Scale,”Nature 529(2016):910.81“The Chinese Communist Party Central Committee and the State

222、Council Issue Outline of the Healthy China 2030 Plan”(中共中央 國務院印發“健康中國2030”規劃綱要),Xinhua News Agency,October 25,2016,https:/perma.cc/8576-U8HX.82 CSET translation of“State Council Notice on the Publication of the National 13th Five-Year Plan for S&T Innovation”(國務院關于印發”十三五”國家科技創新規劃的通知),Central Peoples

223、 Government of the Peoples Republic of China,July 28,2016,https:/cset.georgetown.edu/research/state-council-notice-on-the-publication-of-the-national-13th-five-year-plan-for-st-innovation.83 CSET translation of“National 13th Five-Year Plan for the Development of Strategic Emerging Industries”(國務院關于印

224、發”十三五”國家戰略性新興產業發展規劃的通知),Central Peoples Government of the Peoples Republic of China,November 29,2016,https:/cset.georgetown.edu/research/national-13th-five-year-plan-for-the-development-of-strategic-emerging-industries/.84 CSET translation of“Notice of the State Council on the Publication of Made in

225、 China 2025”(國務院關于印發中國制造2025的通知),Central Peoples Government of the Peoples Republic of China,May 8,2015,https:/cset.georgetown.edu/publication/notice-of-the-state-council-on-the-publication-of-made-in-china-2025.85 International Telecommunication Union(ITU)translation of“The National Medium-and Long-Term Program for Science and Technology Development(2006-2020):An Outline”(國家中長期科學和技術發展規劃綱要（20062020年）),State Council,2006,https:/www.itu.int/en/ITU-D/Cybersecurity/Documents/National_Strategies_Repository/China_2006.pdf.