SAMURAI-Pro

  • Research type

    Research Study

  • Full title

    Systematic Assessment for the Medical Utility of Radiological and diagnostic Artificial Intelligence – Prospective Evaluation

  • IRAS ID

    357391

  • Contact name

    Alexander Thomas Novak

  • Contact email

    alex.novak@ouh.nhs.uk

  • Sponsor organisation

    Oxford University Hospital Trust

  • Duration of Study in the UK

    5 years, 8 months, 17 days

  • Research summary

    Doctors, nurses, and other healthcare professionals use many different tests or medical images to help diagnose illnesses, but they sometimes make mistakes when interpreting results which may lead to delayed or incorrect treatment. Many Artificial Intelligence (AI) tools have been developed to help clinicians spot illnesses more accurately and efficiently by alerting them to different abnormalities shown in the test results. Many of these tools are legally approved, and already used in hospitals worldwide. However, while they may have performed well in artificial studies, we do not know how well they work in real NHS hospitals and clinics.
    This trial will test different AI tools in a series of studies to see how they help clinicians in their daily work, and how this affects patients, staff and the NHS as a whole. We will collect information generated as part of normal clinical practice on how different AI tools help improve diagnosis and patient care, and compare the results with the algorithm switched on and off for fixed periods of time. The results will help decide how AI can be best used in real-life healthcare settings. We will look to include many studies in this trial using the same approach, which will allow us to test lots of different AI tools very quickly and efficiently.

    The first study in the trial will test whether using AI software to help clinicians detect broken bones on X rays can improve patient care in emergency departments (A&E) and minor injury units (MIU). Over the study period, AI software will be switched on and off monthly for 6 months, and we will collect data on what happens to patients during these periods, as well as measuring the accuracy of the AI tool, and how staff found the experience of using them.

  • REC name

    South Central - Hampshire A Research Ethics Committee

  • REC reference

    25/SC/0252

  • Date of REC Opinion

    9 Sep 2025

  • REC opinion

    Further Information Favourable Opinion