iCAIRD: the Industrial Centre for Artificial Intelligence Research in Digital Diagnostics
Author: Andrew Brooks
Posted: 26 Sep 2019 | 13:36
The iCAIRD project is working to establish a world-class centre of excellence in the application of artificial intelligence to digital diagnostics. The intention is that iCAIRD will allow clinicians, health planners and industry to work together, enabling research-active clinicians to collaborate with innovative SMEs to better inform clinical questions, and ultimately to solve healthcare challenges more quickly and efficiently.
iCAIRD was announced last November as one of five successful bids to the UK Government’s Industrial Strategy Challenge Fund (ISCF). It brings together a pan-Scotland collaboration of 15 partners from across industry, the NHS, and academia. Industry leadership will be provided by Canon Medical Research (radiology) and Royal Philips (digital pathology). iCAIRD has been awarded £10m by Innovate UK, while partner companies Canon Medical Research Europe Ltd and Royal Philips will provide £5M of additional supportive funding, along with seven actively-engaged SMEs.
The key iCAIRD priorities are to:
- Create infrastructure to develop and apply AI in digital diagnostics, pathology and radiology.
- Fast-track the digitisation of Scottish NHS pathology data to create the largest fully digital pathology laboratory network in Europe.
- Work with partner Canon to develop a network of Safe-Haven Artificial Intelligence Platforms (SHAIP), thus allowing crucially important research while ensuring protection of patients’ personal data.
- Work with partner Philips to establish a Health Data Research UK (HDRUK) national pathology image archive of anonymised disease cases within the National Data Safe Haven that can be used to train computers to augment and improve on current practice
Some of the first applications will be to validate AI in stroke medicine, chest X-ray triage, and mammogram interpretation, and apply AI to colon cancer data and gynaecological pathology.
One of EPCC’s roles within iCAIRD is hosting the National Data Safe Haven and the national archive of digital pathology images. The archive is expected to grow to a very substantial amount, in the order of many hundreds of TB.
We will work with Glencoe Software to deploy their OMERO Plus database system, which is ideally suited to managing such a vast collection. Users will be able to import, catalogue and view pathology images, add metadata, draw figures, collaborate with other users and groups, and analyse in a variety of ways. The other role for EPCC is to provide the technology for the application of AI to digital diagnostics, pathology and radiology. This will bring together our expertise in image archives, high performance computing and data science.
iCAIRD is centred at the Imaging Centre of Excellence (ICE), part of the University of Glasgow’s Clinical Innovation Zone (CIZ) at Queen Elizabeth University Hospital. The academic partners are the universities of Edinburgh, St Andrews, Aberdeen and Glasgow. It will integrate with Health Data Research UK (HDRUK) and the national Picture Archiving Communication System (PACS), working closely with partners across Scotland and beyond.
Author
Andrew Brooks, EPCC