Recipient organisationUniversity of GlasgowSource-published name: University of Glasgow
Funding£11.6M
PeriodFeb 2019 — Jun 2023
In plain English
AI plain-English summary
Scotland’s largest hospital is wiring its pathology department—the biggest in Europe—entirely with digital microscopes, and linking that data to a national AI network for faster diagnosis. This matters because the NHS collects vast amounts of medical images—X-rays, scans, tissue slides—but most sit in isolated systems, unread by computers. Without shared, secure data, artificial intelligence cannot be trained to spot disease earlier. iCAIRD solves that by building “Safe-Haven AI Platforms” within existing NHS data centres in Glasgow and Aberdeen, so patient data never leaves the health service while companies and researchers can still develop diagnostic tools. If successful, the project will transform how common conditions are caught. Canon and Philips have already put in £5 million. Small companies are lined up to work on stroke medicine, chest X-ray reading, gynaecological pathology, mammogram analysis, and colon cancer prediction. The same infrastructure could eventually scale nationwide, turning routine scans into early-warning systems without compromising privacy.
View original technical description
"iCAIRD will be a Scottish centre of excellence, centred at the Queen Elizabeth University Hospital in Glasgow but with sites across Scotland, to enable joined up academic and commercial technology development and research, locally and nationally. Making use of the capability of modern computers to process the large amounts of data gathered in healthcare clinics iCAIRD will enable better and earlier diagnosis of conditions, sooner treatment and better outcomes for patients. iCAIRD, draws together Scotland's expertise in clinical and academic digital radiology and pathology, advanced data storage, data governance and access, inter-operable databases and will develop a vibrant community using multiscale data sources to develop clinically relevant and commercially practicable solutions. It adds to and complements collaboration with other national initiatives led by NSS. With Canon, we will build a network of ""Safe-Haven Artificial Intelligence Platforms"" (SHAIP) within existing NHS data 'safe-havens'. We will do this in Glasgow and then Aberdeen, eventually scaling nationwide. This will support a distributed approach to training AI whilst ensuring, identifiable patient data stays in the NHS. The regional nature also enables small companies (SMEs) to work closely with doctors at a local level. We will complete the conversion of the Glasgow Pathology Department, the largest in Europe, from conventional microscopes to digital technology. By linking with the Edinburgh Parallel Computing Centre we will establish a secure, National Pathology Research Image Database with guidance from Philips. This will receive and store images from different sites, even when obtained using different equipment, ensuring that data is accessible as widely as possible, maximising benefit. Canon Medical and Philips are providing £5M of additional funding to support iCAIRD. Once operational, we will seek to create further projects with the aim of leveraging further industry support. However, several SMEs are already involved in this work and will access data and develop solutions for stroke medicine; chest x-ray reading; gynaecological pathology; mammogram analysis; and, improving diagnosis and predictions for colon cancer. iCAIRD will adopt an ethical framework based on current best practice within HDRUK, ensuring patient data are securely held within the NHS whilst enabling companies, researchers and clinicians to work together to develop more effective and efficient diagnostic tools for patients. We will support SMEs to engage with healthcare challenges through the development of an Accelerator Programme providing access to regulatory and route to market expertise from our partner companies, driving economic growth as well as patient benefits."
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know