Deep Learning to Estimate Cardiovascular Risk From Chest Radiographs : A Risk Prediction Study
An erratum to this article was published (Ann Intern Med, Jan 2025 ↗)
Performance · 10-year MACE, outpatients with unknown ASCVD risk (n=8869)Result — go to the resultAdjusted hazard ratio 1.73 95% CI 1.47–2.03Authors' conclusion
On the basis of a single CXR, CXR CVD-Risk predicts 10-year MACE beyond the clinical standard and may help identify individuals at high risk whose ASCVD risk score cannot be calculated because of missing data.
Study context
- Population
- Outpatients potentially eligible for primary CV prevention8,869 unknown + 2132 known atherosclerotic cardiovascular disease (ASCVD) risk
- Comparison
- CXR CVD-Risk vs ASCVD risk score
- Outcome
- 10-year MACE predicted by CXR CVD-Risk vs the ASCVD risk score
- Follow-up
- 10 years
Methodological detailsretrospective, external validation, electronic medical records, Model developed using data from a cancer screening trial, Externally validated
- Retrospective
- External validation
- Electronic medical records
- Model developed using data from a cancer screening trial
- Externally validated
Performance
As reported in the abstract
10-year MACE, outpatients with unknown ASCVD risk (n=8869)
Adjusted hazard ratio
1.73
95% CI 1.47–2.03
CXR CVD-Risk ≥7.5%vs CXR CVD-Risk <7.5%
adjusted HR 1.73 (95% CI 1.47–2.03)
Further findings
10-year major adverse cardiovascular events (MACE) beyond ASCVD risk score, outpatients with known ASCVD risk (n=2132)
| Adjusted HR (CI) | ||
|---|---|---|
| CXR CVD-RiskBeyond ASCVD risk score | 1.88 (1.24–2.85) | |
| 0.20.5125Adjusted hazard ratio (log scale) |
Original publication
Deep Learning to Estimate Cardiovascular Risk From Chest Radiographs : A Risk Prediction Study. Ann Intern Med. Published 26 March 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →