Researchers from the University of Aberdeen are to take part in a project looking at how AI can support clinical decision-making in emergency care.

The project, led by Queen Mary University of London, has secured a £1.8 million contract to support a clinical trial using the Artificial Intelligence in Trauma Risk Prediction System (AI-TRiPS), a support tool designed to assist in time-critical medical decision making for severely injured patients. 

The study will see AI-TRiPS deployed across the London Trauma System, the largest integrated trauma network in the world, serving over 10 million people, and will involve specialists from London’s four major trauma centres, as well as doctors and paramedics from London’s ambulance services.

Aberdeen University’s Clinical Trial Unit will work to design the randomised controlled trial and then evaluate the impact of the decision support system for patients, doctors, and frontline emergency workers.

The University said that, if successful, the initiative could revolutionise trauma care, by helping medical staff assess the risks of life-threatening complications such as severe blood loss, and support them in taking action to improve outcomes. 

Marion Campbell, Professor of Health Services Research at the University of Aberdeen, said: “This project will provide crucial evidence on how innovative AI technology might enhance trauma care and we are delighted to be part of the team conducting this exciting research.”  

The AI algorithms used by AI-TRiPS have been developed by trauma surgeons, military experts, and computer scientists, integrating cutting-edge trauma research, registry data, and clinical expertise.

According to the team, the system is designed to be user-friendly, giving clear and accessible insights to doctors making time sensitive decisions, offering evidence-based predictions about the risks faced by critically injured patients and guidance on how best to manage these on arrival in hospital.

The system explains the reasoning behind its predictions, providing transparency with an ‘open box’ design that can be easily understood and explored by clinical users. 

“This is a pioneering step forward in trauma care,” said Colonel Nigel Tai, lead investigator for the study at Queen Mary University.

“The AI-powered tools we want to evaluate have been co-designed by trauma clinicians, working hand-in-glove with computer scientists. Whilst many AI applications have been developed, few are trialled, meaning that doctors and patients can’t make good judgements about safety and efficacy, and developers lack feedback.

“We think that victims of major trauma – civilian and military – stand to benefit from new technologies, designed to give clinical teams the right information about their patients when they need it most.”

Recommended reading

• Scotland’s Emergency Workers To Get AI-Powered Trauma App
• Digital Inclusion ‘Vital First Step’ for Better Care, Says InnoScot Health
• Work Under Way for Scotland’s First ‘Fully Digital Hospital’

This project is just the latest example of AI tools gaining ground within emergency healthcare. In October, for instance, researchers from the University of Edinburgh announced they were developing a tool using machine learning to help predict which individuals are at high risk of urgent hospital care.

Meanwhile, at the Queen Elizabeth University Hospital in Glasgow staff are trialling artificial intelligence to help improve turnaround times for CT scans and reduce A&E pressures, and in September a study led by Aberdeen’s Robert Gordon University developed AI techniques to identify patients’ risk of developing a stroke.