Stroke patients in England are receiving life-changing treatment more than an hour earlier thanks to a “fast-track” AI tool, a major national study finds. Published in The Lancet Digital Health, the analysis shows about 15,000 patients directly benefited after their scans were reviewed by the technology.
The Brainomix 360 Stroke imaging tool, now rolled out across a network of more than 70 hospitals, analyses CT scans in real time to identify features of a major stroke within minutes. By helping clinicians spot large clots quickly, the tool speeds up decisions and enables faster transfer to specialist stroke centres where thrombectomy — a minimally invasive clot-removing procedure — can be performed. Faster treatment substantially increases the chance of regaining independence after a major stroke.
At hospitals using the tool, thrombectomy rates doubled at participating sites (from 2.3% to 4.6%), compared with smaller increases at hospitals not using the technology (1.6% to 2.6%). The study found the biggest improvements at hospitals without on-site neuroradiology expertise, where rapid interpretation is most critical. At primary stroke centres, AI use was associated with a 64-minute reduction in door-in-door-out time.
Time is critical: every 20-minute delay in thrombectomy reduces the chance of full recovery by around 1%. With roughly 80,000 strokes in England each year, faster diagnosis and decision support from AI could help thousands more patients reach specialist treatment in time to improve outcomes.
Dr David Hargroves, NHS National Clinical Director for Stroke and co-author of the study, said the publication provides robust, real-world evidence that stroke AI imaging delivers faster decision-making and better care, supporting clinicians to give more patients life- and disability-saving treatments and a better chance of returning to independent living.
Patient story: Jean Hines, 83, was taken to Royal Berkshire Hospital after collapsing. A scan supported by Brainomix identified a major stroke and the need for urgent specialist treatment. Within 25 minutes of arrival she was transferred by blue-light ambulance to the John Radcliffe Hospital in Oxford and underwent a successful mechanical thrombectomy. Jean attributes her good physical recovery to the speed of treatment and said she felt “incredibly lucky” that rapid care prevented serious disability.
Key facts from the study and rollout:
– 452,952 stroke patients included in the national audit dataset
– 71,017 patients treated at the 26 evaluation hospitals in the patient-level analysis
– After AI implementation (Jan 2022 onwards), 15,377 patients had their scans reviewed using AI
– At evaluation sites, thrombectomy rates doubled from 2.3% to 4.6%
– At primary stroke centres, AI use associated with a 64-minute reduction in door-in-door-out time
– Patients reviewed with AI were more likely to receive thrombectomy and intravenous thrombolysis
– Patients reviewed with AI were more likely to achieve a favourable functional outcome at discharge, with no increase in in-hospital mortality
– Study: “Artificial intelligence imaging decision support for acute stroke treatment in England: a prospective observational study,” Lancet Digital Health, 2025
– AI decision-support tools were rolled out to all stroke centres in England in Summer 2024 as part of the national optimal stroke imaging pathway
