GUIDE-IT Meeting
17 September 2026
On 17 September 2026, the GUIDE-IT community met at the Riley Auditorium, Clare College, Cambridge. More than 70 participants from over 30 institutions in 11 countries came together, including clinical trialists, radiologists, cardiologists, AI developers, industry partners, funders and journal editors.
The meeting brought together leading cardiac CT and prostate MRI trial groups to discuss how imaging data from randomised controlled trials can be shared and reused responsibly. Sessions covered the current state of the GUIDE-IT platform, the onboarding of new trials, opportunities for independent AI benchmarking, and the governance and funding needed to sustain trial data sharing in the long term.
We thank all participants and speakers for their contributions and look forward to continuing the collaboration.
Sessions
The GUIDE-IT Meeting consisted of four sessions, focusing on advances in medical imaging and data sharing in particular.
-
Session 1: Improving Patient Outcomes through Data Sharing - The Example of Coronary CT
-
Session 2: Infrastructure Perspectives on Data Sharing
-
Session 3: Improving Patient Outcomes through Data Sharing - The Example of Prostate MRI
-
Session 4: Industry Perspectives on AI in Radiology
Each of the sessions was followed by a discussion between presenters and the moderators Professor Giganti (University College London) and Professor Dewey (University of Cambridge).
The learning from this meeting will serve as an input to advance from prototype stage to public launch.
Randomised Controlled Imaging Trials
Randomised controlled imaging trials test how imaging should be used in patient care. Patients are randomly assigned to different imaging strategies, for example CT versus invasive angiography, or MRI before versus after biopsy. The outcomes are then followed over time. These trials provide the strongest evidence for which imaging test benefits patients, and they directly shape clinical guidelines.
Between 2014 and 2023, 375 randomised controlled imaging trials were published in ten leading medical and radiological journals. Running them took 578 research sites across 49 countries, from large academic centres to small district hospitals, on every inhabited continent. Together these trials cover a wide range of diseases and imaging methods and hold outcome data from many thousands of patients. Most are publicly funded, yet their imaging data are rarely available once the results are published, so each trial tends to end as a closed study rather than a shared resource.
GUIDE-IT brings this community together. By combining imaging and outcome data from many trials in a trusted research environment, we want to answer questions no single trial can answer alone, validate AI tools against real patient outcomes, and give every trial a lasting second life.
