Overview

A research initiative from the University of Exeter is leveraging AI and machine learning to customize stroke care for patients, aiming to decrease the likelihood of future disability.

Stroke Statistics

• Stroke is a major cause of death and disability.
• Over 100,000 individuals are hospitalized in the UK annually due to strokes.

About the SAMuEL Project

The Stroke Audit Machine Learning (SAMuEL) project utilizes AI to assist healthcare professionals in identifying patients who are most likely to benefit from thrombolysis, a clot-busting treatment that can significantly reduce disability when administered promptly.

Key Insights from Professor Martin James

Professor Martin James, a consultant stroke physician and honorary clinical professor at the University of Exeter Medical School, stated:

“SAMuEL analysis incorporates nationwide data from a quarter of a million stroke cases. By utilizing this data, we can provide each hospital with a customized target for thrombolysis. When teams have used this as a benchmark, they’ve been able to treat more patients effectively.”

He added, “Stroke has a life-changing impact, so it’s encouraging to see how research like this can lead to more personalized, faster treatment and improved outcomes for patients and their families.”

Partnerships and Development

This project is a collaboration with the Royal Devon University NHS Foundation Trust and the National Institute for Health and Care Research Applied Research Collaboration South West Peninsula (PenARC).

During the SAMuEL 2 phase, which took place from 2022 to 2024, a tool was created to enhance the understanding of how clot-busting medications are utilized in hospitals, ensuring that more patients receive optimal treatment swiftly.

Significance of the Research

The University of Exeter claims this is the first instance of integrating AI into a national stroke audit, which is expected to improve the targeting of thrombolysis treatment for local populations.

• Thrombolysis has been administered to about 11% of stroke patients in recent years.
• Over 1,000 patients per year in the South West have received this treatment.

Challenges and Future Directions

It is important to note that thrombolysis is not appropriate for every patient and is only effective when given shortly after a stroke. The frequency and speed of thrombolysis administration can vary significantly across different regions.

Building on prior research, the team at the University of Exeter employed computer models to analyze the variability in thrombolysis use among hospitals. They also aimed to predict patient outcomes and identify key patient characteristics that influence recovery after a stroke, both with and without thrombolysis.

Future Research

The methods developed in this study for analyzing national stroke audit data may be applicable to other national clinical audits, such as those related to maternity care.

Future phases of research, including SAMuEL 3, which commenced in April 2025 and will last for two years, will expand the focus to other stroke treatments, such as thrombectomy (mechanical clot removal) and will analyze individual brain scans to better determine who will benefit from thrombolysis and who may be at risk of harm.