AI for Stroke: What the Data Show
Researchers evaluated artificial intelligence stroke imaging across England's health service stroke network.
-
By
-
Kathryn Wighton
-
January 14, 2026
-
Clinical Scorecard: AI for Stroke: What the Data Show
At a Glance
| Category | Detail |
|---|
| Condition | Acute Stroke |
| Key Mechanisms | Implementation of AI imaging software improves endovascular thrombectomy rates and reduces interhospital transfer times. |
| Target Population | Patients aged 16 years and older admitted to NHS hospitals with a primary diagnosis of stroke. |
| Care Setting | NHS hospitals in England. |
Key Highlights
- Endovascular thrombectomy rates doubled at hospitals using AI imaging software.
- AI support reduced median door-in door-out time by 64 minutes.
- 69% of non-evaluation sites adopted AI software during the study period.
- AI use associated with higher rates of intravenous thrombolysis.
- No association found between AI use and in-hospital mortality.
Guideline-Based Recommendations
Diagnosis
- Utilize AI imaging support for acute stroke diagnosis.
Management
- Implement AI software to increase endovascular thrombectomy rates.
Monitoring & Follow-up
- Track functional outcomes using the modified Rankin Scale.
Risks
- Consider limitations of observational studies in causal inference.
Patient & Prescribing Data
Patients with ischemic stroke treated at evaluation sites.
AI-supported imaging interpretation increases likelihood of endovascular thrombectomy.
Clinical Best Practices
- Adopt AI imaging software in primary stroke centers.
- Monitor door-in door-out times to optimize transfer efficiency.
References
