Machine learning of clinical variables and coronary artery calcium scoring for the prediction of obstructive coronary artery disease on coronary computed tomography angiography: analysis from the CONFIRM registry
Authors' conclusion
A ML model incorporating clinical features in addition to CACS can accurately estimate the pretest likelihood of obstructive CAD on CCTA.
Study context
- Population
- Patients undergoing ≥64 detector row coronary computed tomography angiography (CCTA) for suspected or established coronary artery disease (CAD), CONFIRM registry35,281 screened
- Participants
- 13,054 patients
- Comparison
- Machine learning (ML) + coronary artery calcium score (CACS) vs ML alone, CAD consortium clinical score and updated Diamond-Forrester (UDF) score
- Outcome
- Prediction of obstructive CAD (≥50% stenosis) on CCTA
Methodological detailsregistry-based, machine learning (XGBoost), 80% training set with 10-fold cross-validation, 20% test set
- Registry-based
- Machine learning (XGBoost)
- 80% training set with 10-fold cross-validation, 20% test set
Performance
As reported in the abstract
Obstructive CAD on CCTA
| Test or model | AUC |
|---|---|
| ML + CACS | 0.881 |
| ML | 0.773 |
| CAD consortium clinical score | 0.734 |
| CAD consortium clinical score + CACS | 0.866 |
| UDF | 0.682 |
AUC
ML + CACS0.881
ML0.773
CAD consortium clinical score0.734
CAD consortium clinical score + CACS0.866
UDF0.682
- ML + CACS best performance vs all comparators
- P < 0.05 for all comparisons
Further findings
Prevalence of obstructive CAD (≥50% stenosis)
- Patients with obstructive CAD
- 2380 (18.2%)
Feature importance
- CACS, age and gender were the highest ranking features
Original publication
Machine learning of clinical variables and coronary artery calcium scoring for the prediction of obstructive coronary artery disease on coronary computed tomography angiography: analysis from the CONFIRM registry. Eur Heart J. Published 12 September 2019. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →