EXAM [COVID-19]

  • Research type

    Research Study

  • Full title

    EMR CXR AI Model

  • IRAS ID

    287715

  • Contact name

    Quanzheng Li

  • Contact email

    Li.Quanzheng@mgh.harvard.edu

  • Sponsor organisation

    Massachusetts General Hospital & Brigham Women's Hospital Center for Clinical Data Science (CCDS)

  • Duration of Study in the UK

    0 years, 6 months, 1 days

  • Research summary

    The Covid-19 pandemic is caused by the virus SARS-CoV-2 (COVID). which is a respiratory virus. Patients with the disease commonly experience symptoms of a dry cough, fever, fatigue and shortness of breath. Resulting in some patients presenting to hospital and requiring treatment, such as oxygen therapy. Data from patients with and without COVID has been collected as part of research Cambridge University Hospitals NHS Foundation Trust. This data includes information stored in electronic medical records and imaging data such as chest-x-rays. The data is essential for the further understanding of this disease as well as the development of new tools to help in targeting resources to treat patients with the disease. One such tool is a computer algorithm that will predict the likelihood of a patient requiring oxygen therapy. This tool will help with allocating resources through the prediction of which patients will possibly require oxygen therapy. \n\nThis data has been de-identified (so that all information that could be used to identify an individual is removed) and transferred to a secure data store for researchers to use at the University of Cambridge. \n\nFor this project controlled access to the data at Cambridge will be provided for the training of the specialised tool, no data will leave site and once trained only the tool will be returned to the host organisation. This process of training will take place at 20 other sites around the world.Collaboration between industry, academia and clinicians is essential to this work being conducted and new innovative ways such as federated learning allow hospital sites to have more control over their data and create models that are generalisable to the world wide population.[Study relying on COPI notice]

  • REC name

    South West - Central Bristol Research Ethics Committee

  • REC reference

    20/SW/0140

  • Date of REC Opinion

    2 Sep 2020

  • REC opinion

    Favourable Opinion