《Nature,5月22日,Machine Learning for COVID-19 needs global collaboration and data-sharing》

  • 来源专题:COVID-19科研动态监测
  • 编译者: xuwenwhlib
  • 发布时间:2020-05-23
  • Machine Learning for COVID-19 needs global collaboration and data-sharing

    Nathan Peiffer-Smadja, Redwan Maatoug, François-Xavier Lescure, Eric D’Ortenzio, Joëlle Pineau & Jean-Rémi King

    Nature Machine Intelligence (2020)

    The COVID-19 pandemic poses a historical challenge to society. The profusion of data requires machine learning to improve and accelerate COVID-19 diagnosis, prognosis and treatment. However, a global and open approach is necessary to avoid pitfalls in these applications.

    On 31 December 2019, the first cases of a viral pneumonia with unknown aetiology were reported in the city of Wuhan, China. In the following weeks, the Chinese authorities and the World Health Organization (WHO) announced the discovery of a novel coronavirus and its associated disease: SARS-CoV-2 and COVID-19, respectively. On 21 April 2020, the number of cases of COVID-19 exceeded 2.4 million and the death toll exceeded 170,000 worldwide1. The outbreak of COVID-19 represents a major and urgent threat to global health. While the unprecedented speed of the COVID-19 spread partly finds its roots in our increasingly globalized society, the global sharing of scientific data also offers a promising tool to fight the disease. In the past four months, more than 12,400 articles have been published2 and scientific data collected from thousands of patients have been released3. The majority of these studies follow the standard scientific method: that is, investigate a few hypotheses at a time on a controlled sample. While undeniably successful, this standard method suffers from two well-known challenges, both critical to our pandemic situation: (1) it requires considerable expertise and human input and (2) it only considers a handful hypotheses at a time. Machine learning (ML) has been used to meet these challenges in various pathologies4,5, including infectious diseases6. Herein, we describe two areas where ML could supplement standard statistical methods in the COVID-19 pandemic, discuss the practical challenges that such a ML approach entails, and advocate for a global collaboration and data-sharing.

  • 原文来源:https://www.nature.com/articles/s42256-020-0181-6
相关报告
  • 《Nature,5月5日,COVID-19 pandemic: the effects of quarantine on cardiovascular risk》

    • 来源专题:COVID-19科研动态监测
    • 编译者:zhangmin
    • 发布时间:2020-05-06
    • COVID-19 pandemic: the effects of quarantine on cardiovascular risk Anna Vittoria Mattioli, Matteo Ballerini Puviani, Milena Nasi & Alberto Farinetti European Journal of Clinical Nutrition (2020) Abstract COVID-19 is causing a global pandemic with a high number of deaths and infected people. To contain the diffusion of COVID-19 virus, Governments have enforced restrictions on outdoor activities or even collective quarantine on the population. One important consequence of quarantine is a change in lifestyle: reduced physical activity and unhealthy diet. 2019 guidelines for primary prevention of cardiovascular disease indicate that “Adults should engage in at least 150 minute per week of accumulated moderate-intensity or 75 minute per week of vigorous-intensity aerobic physical activity (or an equivalent combination of moderate and vigorous activity) to reduce ASCVD risk.” During quarantine, strategies to further increase home-based physical activity and to follow a healthy diet should be implemented. Quarantine carries some long-term effects on cardiovascular disease, mainly related to unhealthy lifestyle and anxiety. Following quarantine a global action supporting healthy diet and physical activity is mandatory to encourage people to return to good lifestyle.
  • 《5月22日_COVID-19的机器学习需要全球协作和数据共享》

    • 来源专题:COVID-19科研动态监测
    • 编译者:zhangmin
    • 发布时间:2020-05-24
    • 1.时间:2020年5月22日 2.机构或团队:巴黎第七大学UDR医学院、法国巴黎皮蒂埃-萨尔佩特里埃尔医院、法国Bichat-Claude-Bernard医院、法国卫生和医学研究所、加拿大麦吉尔大学、加拿大Facebook AI 实验室等 3.事件概要: Nature Machine Intelligence 期刊于5月22日发表了法国巴黎第七大学UDR医学院等机构的评论文章“Machine Learning for COVID-19 needs global collaboration and data-sharing”。文章称,COVID-19大流行给社会带来了历史性挑战。大量的数据需要机器学习(ML)来改善和加速COVID-19的诊断、预后和治疗。但是,必须采用全局和开放的方法来避免这些应用程序中的陷阱。 文章表示,尽管COVID-19的空前传播速度在一定程度上源于日益全球化的社会,但全球科学数据共享也为抗击这种疾病提供了有希望的工具。在过去的四个月中,已经发表了超过12,400篇COVID-19相关的文章,这些研究大多数遵循标准的科学方法,即一次在受控样本上研究一些假设。尽管这种方法无可否认是成功的,但它却面临着两个众所周知的挑战,即它需要大量的专业知识和人力投入以及一次只考虑少数几个假设。本文中研究人员描述了ML可以补充COVID-19大流行中的标准统计方法的两个领域,讨论了这种ML方法所带来的实际挑战,并倡导全球协作和数据共享。 文章表示,ML可以减轻医学专家的工作量。ML技术可以减少进行自动化分析所需的时间,并允许人工智能从业人员为临床医生提供支持。但是,目前仅限于相对较小的队列使用ML医学图像来诊断或预后COVID-19。因此,这些研究无法很好地控制算法可能从胸部图像中检测到的许多混杂因素(例如年龄、肥胖)。一种有前途的策略是从相似图像的更大数据集中预训练ML模型,从而学习要计算的通用特征,然后可以使用这些特征来促进从COVID-19图像进行训练。此外,ML可以加快治疗筛选。文章指出,标准方法一次只考虑少数假设。例如,在已注册的用于识别COVID-19的治疗方法的1200多个临床试验中,大多数集中于一种或多种药物的独特选择,这些药物是根据相关性的不同而手工选择的。ML可以通过同时考虑几种潜在的抗病毒药物,依靠DNA序列和/或蛋白质结构(包括SARS-CoV-2的潜在药物结合位点)来预测药物与病毒之间的相互作用,从而拓宽此类筛选和选择过程。文章表示,尽管如此,不应忽略基于ML的治疗研究的挑战。首先,ML无法加速基础生物学,甚至蛋白质折叠的预测仍然是一个非常困难的问题。对于疫苗,有一个必要的等待期。其次,一个主要的伦理问题是绕过适当的临床试验的诱惑,机器学习算法可能会大大增加这种风险。 文章表示,总的来说,机器学习有望补充而不是取代用于诊断、预后和治疗的标准方法。目前有两个主要挑战限制了机器学习的潜在影响。首先,医疗保健专业人员必须意识到,像人类一样,机器学习很容易受到系统性偏差的影响(例如,扫描设备、患者年龄等)。其次,缺乏大型医疗、临床、影像和基因公共资源库,导致每个机构都在自己的小型数据集上本地开发自己的分析管道,这极大地限制了结果的可推广性。文章指出,国际严重急性呼吸系统和新兴感染联合会(ISARIC)计划旨在提供有关COVID-19患者的大型共享临床数据库。其他机构也签署了数据共享协议,以确保广泛,快速地共享数据,并且可以为新的假设提供依据,但是这仍然是零碎的,因此难以充分利用数据共享大流行期间每天产生的数据。标准和ML模型的质量不仅直接取决于此类数据库的大小、质量和代表性,而且对于支持跨不同国家和医疗机构类型进行有效干预至关重要。 4.附件: 原文链接:https://www.nature.com/articles/s42256-020-0181-6