AI Enhances Stroke Diagnosis and Treatment Timing
A new AI-powered brain scan analysis tool can accurately determine when a stroke occurred, improving emergency treatment decisions. Developed by Imperial College London, Technical University of Munich, and Edinburgh University, the software is twice as accurate as traditional visual assessments by doctors.
Why it matters:
- Stroke treatment is time-sensitive—within 4.5 hours, patients are eligible for clot-busting drugs and surgery.
- After six hours, treatment decisions become more complex, making precise timing critical.
- Some strokes occur while patients are asleep, making onset time difficult to determine.
How it works:
- The AI was trained on 800 brain scans and tested on 2,000 patients, learning to analyse lesion texture and progression.
- It estimates both the chronological start time and the biological reversibility of the stroke.
- Fully automated, it could increase appropriate stroke treatments by 50%.
The researchers aim to integrate this technology into the NHS, enhancing stroke care across hospitals.