AI Summary
What this is: An AI tool called Brainomix now reads brain scans in NHS hospitals across England, cutting the time between a stroke patient arriving and getting treatment.
Who it’s for: Anyone with an older relative, a family history of stroke, or a general interest in how AI is changing hospital care.
The key idea: Speed decides stroke outcomes, and most of the damage happens during the wait for a diagnosis. AI shortens that wait by getting the scan read and the treatment decision made faster.
Skip if: You want a general explainer on how AI works in medicine. Try AI in Medical Imaging instead.
A stroke kills about two million brain cells a minute if it goes untreated. For decades, the bottleneck wasn’t medicine, it was time: a scan has to be taken, read by a specialist, and acted on, often in the middle of the night, often at a hospital with no stroke expert on call. Across England, an AI tool called Brainomix now does the reading in minutes instead of hours, and the results are showing up in survival data, not just marketing copy.
What is this AI stroke tool, and how does it work?
Brainomix, spun out of stroke-imaging research at Oxford University, analyzes a patient’s brain scan the moment it’s taken and flags whether they’re a candidate for AI-assisted emergency treatment. The doctor still makes every call. What changes is the wait: a first read that used to require a specialist to be physically available, day or night, now takes minutes, so a smaller hospital without an on-site stroke expert can act immediately instead of waiting for a consultant or transferring the patient elsewhere first.
That single change, cutting the wait for a first read, is the whole story. Everything downstream (deciding on clot-busting drugs, deciding on a thrombectomy to physically remove the clot) depends on how fast that first decision gets made.
The numbers: faster treatment, more full recoveries
At Royal Berkshire Hospital, one of the earliest sites to roll out the tool, the share of stroke patients who recovered enough to live independently roughly tripled, from 16% to 48%, and the time between a patient arriving and treatment starting dropped by 61 minutes, according to a study published in Frontiers in Neurology and reported alongside the NHS rollout announcement.
A larger follow-up study in The Lancet Digital Health, covering close to 453,000 stroke patients across all 107 NHS hospitals in England between 2019 and 2023, found that thrombectomy rates (the rate at which patients got the highest-impact stroke treatment) doubled at the 26 hospitals using Brainomix 360 Stroke, from 2.3% to 4.6%. Hospitals without the tool saw a smaller rise, from 1.6% to 2.6%, in the same period.
| Measure | Before AI triage | After AI triage |
|---|---|---|
| Full-recovery rate (Royal Berkshire Hospital) | 16% | 48% |
| Time to treatment (Royal Berkshire Hospital) | 140 minutes | 79 minutes |
| Thrombectomy rate (26 hospitals, Lancet study) | 2.3% | 4.6% |
| Thrombectomy rate (hospitals without the tool) | 1.6% | 2.6% |
That table is our own read of the two studies side by side; neither the NHS nor the Lancet paper published them together this way, but lining them up shows the pattern clearly: everywhere Brainomix was deployed, both speed and outcomes moved together.
Why every minute matters after a stroke
Every 20-minute delay in starting a thrombectomy lowers a patient’s odds of a full recovery by roughly one percentage point, according to the NHS England writeup of the Lancet study. That sounds small until you multiply it by the hours a patient used to wait for a scan to be read by a human specialist, especially overnight or at a hospital without a stroke unit.
Every minute saved during the initial hospital assessment of people with stroke-like symptoms can dramatically improve a patient’s chance of leaving hospital in good health.
Dr Timothy Ferris, NHS England Director of Transformation
Knowing the warning signs matters just as much as any hospital technology, since none of this helps if a stroke isn’t recognized in the first place. The F.A.S.T. checklist, used across US and UK public health campaigns, covers the four signs to watch for:
- Face: one side of the face droops or feels numb when the person smiles
- Arms: one arm drifts downward when both are raised
- Speech: slurred, garbled, or hard to understand
- Time: if any of the above appear, call emergency services immediately and note when symptoms started
How many hospitals are using this right now?
NHS England reported in December 2025 that Brainomix is deployed at more than 70 NHS hospitals, with roughly 15,000 patients directly benefiting from AI-reviewed scans since the rollout began. The tool started as a pilot funded through the NHS’s AI in Health and Care Award, covering five stroke networks, before expanding nationally.
Could this technology come to the US?
The underlying idea, using AI to triage brain scans for stroke speed rather than diagnosis, is already spreading beyond the NHS. Similar AI stroke-imaging tools operate in some US hospital systems, though nothing at the scale of the NHS’s single-payer, single-dataset rollout. A single health system that can mandate and measure a tool across every hospital in the country at once is what let the Lancet study track 453,000 patients instead of a few hundred, and that scale is the real NHS advantage here, not the AI itself.
What AI still can’t do in a stroke ward
Reading a scan is the whole job. Calling the patient’s family, explaining a diagnosis to someone who is frightened and confused, and making the judgment call on a borderline case where a neurologist’s experience matters more than a probability score all stay with the clinical team. Brainomix speeds up the decision a doctor makes; it doesn’t make that decision for them. Every account of the rollout, including the NHS’s own writeup, describes it as a tool the clinical team uses, not a replacement for one.
What this means if you or someone you love has a stroke
Stroke symptoms move fast, and so should you: sudden face drooping, arm weakness, or slurred speech means calling emergency services immediately, not waiting to see if it passes. The AI tool in this story only helps once someone is already at a hospital that has it. It doesn’t change the first, most important step, which is still getting to a hospital as fast as possible. If you have an older relative or anyone at elevated stroke risk, it’s worth knowing which hospital in your area treats the most stroke cases; specialist stroke units, with or without AI triage, consistently produce better outcomes than general emergency rooms.
Where this fits in the bigger AI-in-medicine picture
Stroke imaging is one example of a broader pattern: AI handling the fast, narrow, repeatable task, freeing up a specialist’s time for the judgment calls that need a human. We’ve covered similar ground in AI in Medical Imaging, and if you’re curious what AI can and can’t safely do with your own health questions, our own experiment in My AI Medical Second Opinion is a useful next read.
Frequently asked questions
Does the AI diagnose the stroke?
No. It flags whether a patient’s scan is a strong candidate for time-sensitive treatment, so the specialist can decide faster. The clinical decision still comes from a doctor.
Is this the same as Claude or ChatGPT reading a scan?
No. Brainomix is a purpose-built medical imaging tool trained specifically on stroke scans, cleared for clinical use, and integrated directly into NHS hospital systems. It isn’t a general-purpose chatbot.
Which hospitals use it?
NHS England reported more than 70 hospitals using Brainomix 360 Stroke as of December 2025, up from an initial five-network pilot. Check with your local hospital trust for its specific stroke-unit capabilities.
Does faster treatment guarantee a full recovery?
No. It significantly improves the odds, doubling thrombectomy rates and roughly tripling full-recovery rates at the earliest pilot site, but stroke severity, the patient’s health going in, and how quickly they reached a hospital in the first place all still matter.
Is a similar tool available in the US?
Some US hospital systems use comparable AI stroke-imaging tools, though there’s no equivalent nationwide, single-system rollout like the NHS’s. Ask your local hospital directly if you want to know what it uses.
Made for regular people, by a real person
A free daily newsletter about using AI in ordinary life. Written by a human, in plain English. No jargon, no paywall.
Free forever. Unsubscribe anytime.
Sources
NHS England: “Life-changing” AI support helping stroke patients get a second chance (December 2025)
Brainomix: new study published in The Lancet Digital Health (PRNewswire, December 2025)
GOV.UK: Artificial intelligence revolutionising NHS stroke care (December 2022)
Oxford University Innovation: Brainomix case study
National Health Executive: AI stroke tool cuts treatment times and doubles chances of recovery
You might also like
AI in Medical Imaging: How AI Reads X-Rays, MRIs, and CT Scans
My AI Medical Second Opinion
Best AI Medical Scribes for 2026
Best AI for Seniors
Best AI Prompts for Doctors
I Built an Agent to Do My Worrying
AI Glossary
Two ways to go further
The AI Prompt Library
1,000+ ready-to-use prompts for Claude, ChatGPT, and Gemini. Stop staring at a blank box.
Get it for $39 →2-Hour Live AI Crash Course
A private, beginner-friendly session across Claude, ChatGPT, Gemini, and the wider landscape.
Book for $125 →