Aortic & peripheral vascular diseaseDiagnostic or prognostic study
AI-based diagnosis of acute aortic syndrome from noncontrast computed tomography (CT)
- Diagnostic accuracy study
- step-wise
- multicenter retrospective study
- large-scale real-world study
- prospective comparative study
- prospective pilot deployment
Study context
- Population
- Patients with suspected acute aortic syndrome (AAS) undergoing noncontrast CT in China
Performance
As reported in the abstract
AAS identification on noncontrast CT
| Test or model | Mean AUC, multicenter retrospective study (n = 20,750) | Sensitivity, real-world study (n = 137,525) | Specificity, real-world study (n = 137,525) |
|---|---|---|---|
| iAorta | 0.958 (95% CI 0.950-0.967) | 0.913-0.942 | 0.991-0.993 |
Further findings
| Outcome | Values |
|---|---|
| AAS correctly identified, prospective pilot deployment in ED (15,584 consecutive patients with acute chest pain) | iAorta 21 out of 22 · Average time to diagnosis (21 AAS-positive) 102.1 (75-133) min |
Time to correct diagnostic pathway, patients with initial false suspicion (prospective comparative study, n = 13,846)
- Average time shortened significantly from 219.7 (115-325) min to 61.6 (43-89) min with iAorta
Original publication
AI-based diagnosis of acute aortic syndrome from noncontrast CT. Nat Med. Published 20 August 2025. doi:10.1038/s41591-025-03916-z
Summarised from the published abstract; every number is checked against it