Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya
Authors' conclusion
In a Kenyan population, machine learning models estimated cardiovascular parameters with comparable or slightly lower accuracy than in the population where they were trained, suggesting model recalibration may be appropriate. This study represents an incremental step toward leveraging machine learning to make early cardiovascular screening more accessible, particularly in resource-limited settings.
Study context
- Population
- Participants at two sites in Kenya with digital retinal fundus photographs, blood pressure (BP), laboratory measures and medical history; models originally trained on UK Biobank dataenrolled
- Participants
- 301 participants
- Comparison
- Diabetes mellitus (DM) model vs Hypertension model
- Outcome
- Performance of UK Biobank-trained models estimating cardiovascular parameters and diagnoses from fundus images
Methodological detailsprospective, non-interventional, external validation
- Prospective
- Non-interventional
- External validation
Performance
As reported in the abstract
| Test or model | AUC | AUC in UK Biobank |
|---|---|---|
| DM model | 0.762 | 0.818 |
| Hypertension model | 0.765 | 0.738 |
AUC
DM model0.762
Hypertension model0.765
AUC in UK Biobank
DM model0.818
Hypertension model0.738
Original publication
Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya. Diabetes Obes Metab. Published 15 April 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →