Artificial intelligence (AI) technologies applied to imaging, such as cancer screening, are among the most advanced uses of AI in healthcare.
Adoption of these technologies has significant potential to both improve care outcomes and reduce costs for the NHS across the UK. The NHS AI Lab wants to ensure that the opportunity of using AI in imaging is shared across the NHS. To do this, the Lab is supporting the development of imaging technology by setting up systems to safely collect and share data. We are providing ways to assess AI imaging software and helping to make it easier to get new technologies into use in hospitals and care settings.
Our work with imaging databases
The NHS AI Lab Imaging programme was developed during the COVID-19 pandemic to support the research and innovation of new technologies in response to the national crisis.
This work has resulted in two major projects of work: a national database of chest images for COVID-19 research (the NCCID) and a second platform for a wider range of medical images (the NMIP).
Project 1: The National COVID-19 Chest Imaging Database (NCCID)
The National COVID-19 Chest Imaging Database (NCCID) is a centralised UK database containing chest X-ray (CXR), magnetic resonance imaging (MRI) and computed tomography (CT) images from hospital patients across the country. A national resource like this is valuable for its diversity and quantity of data, which is being provided to both researchers and technology companies to help train and validate AI models. Having access to datasets within the NCCID will allow AI tools to be developed more quickly and safely.
NHSX has set up the NCCID to support better understanding of the virus and develop technology enabling the best care for patients hospitalised with a severe infection. More than 20 NHS trusts are already contributing data to the National COVID-19 Chest Imaging Database (NCCID).
Project 2: The National AI Medical Imaging Platform (NMIP)
Following the success of the COVID-19 chest imaging database, the NHS AI Lab Imaging team is looking at developing a National AI Medical Imaging Platform (NMIP).
Having completed a discovery exercise that explored whether a national platform would benefit the NHS, and using the learnings from the NCCID project, this platform aims to provide large scale data collection on a national scale, meaning AI developers have access to the kind of high quality data needed to train AI products.
The platform would include large volumes and types of medical imaging for use in testing and developing AI technologies for a range of health and social care needs. Having data from a wide population base and broad geographical spread also avoids unintended bias in the system.
The best AI technologies have the potential to benefit medical professionals, patients, the public and the NHS. In order to get products into use by hospitals and care settings safely and efficiently, the NHS AI Lab is creating a user-friendly route through the testing and regulation that is required.
How to get involved
To contribute chest imaging data to the NCCID, please refer to the information on the NCCID Web Portal.
To access NCCID data for research or investigations into COVID-19, please read our information on how to access the database on the NHSX GitHub website.
Information about the sites that are contributing data toward the database and projects in development can be found on our GitHub.
For further information, please contact the NHS AI Lab Imaging team at email@example.com.
Patient safety and privacy
Patients’ privacy will be safeguarded at every stage of the process and data will only be used for creating tools to support health and social care.
Clinical data and images are de-identified, or made anonymous, at the point of upload. This means that no patient identifiable information will leave hospital sites. The anonymous images become part of data sets that can be used to trial new AI technologies and to validate, or prove the accuracy, of existing ones.
We have worked with the relevant regulatory and advisory bodies to ensure the safety of data and to verify the significant value of the AI Lab’s imaging database projects.
For further information please refer to: