Machine learning score using stress CMR for prediction of all-cause death in patients with suspected or known coronary artery disease a large registry with > 200,000 patient-years of follow-up - 28/12/21
, P. Garot 1, F. Sanguineti 1, T. Unterseeh 1, S. Champagne 1, S. Toupin 3, S. Morisset 4 : Independent biostatistician, T. Hovasse 1, J. Garot 1
Bienvenue sur EM-consulte, la référence des professionnels de santé.
Article gratuit.
Connectez-vous pour en bénéficier!
Résumé |
Background |
Traditional prognostic risk assessment in patients undergoing noninvasive imaging is based upon a limited selection of clinical and imaging findings. Machine learning (ML) can consider a greater number and complexity of variables.
Purpose |
Therefore, we investigated the feasibility and accuracy of ML to predict 10-year all-cause mortality in patients undergoing stress cardiovascular magnetic resonance (CMR), and compared the performance to existing clinical or CMR metrics.
Methods |
Between 2008 and 2018, a retrospective cohort study with a median follow-up of 6.0years (IQR: 5.0–8.0) included all consecutive patients referred for stress CMR. The primary outcome was all-cause death based on the National Death Registry. ML involved automated feature selection by random survival forest, model building with a boosted ensemble algorithm, and 10-fold stratified cross-validation.
Results |
Among the 31,752 consecutive patients (mean age 63.7±12.1years and 65.7% males), 2679 (8.4%) died at 206,453 patient-years of follow-up. Five clinical and 4 CMR parameters were selected by ML including: age, gender, BMI, history of PAD, renal failure, extent of ischemia and LGE, LVEF and LV end-diastolic volume indexed. Regarding the calibration, C-index on the training sets was 0.730 and 0.735 on the test sets, which attests to the very good robustness of ML models. Machine learning exhibited a higher area-under-curve compared with the FRS, ESC-SCORE and QRISK3 score alone for predicting all-cause mortality (ML score: 0.76 vs. FRS: 0.60, ESC-SCORE: 0.63, QRISK3: 0.62; all P<0.001) (Fig. 1).
Conclusion |
Machine learning combining clinical and stress CMR data was found to predict 10-year all-cause mortality significantly better than existing clinical scores alone.
Le texte complet de cet article est disponible en PDF.Plan
Vol 14 - N° 1
P. 48-49 - janvier 2022 Retour au numéro
Article précédent
- Incremental long-term prognostic value of stress CMR above traditional risk factors to predict death a large registry with > 200,000 patient-years of follow-up
- T. Pezel, T. Unterseeh, P. Garot, T. Hovasse, F. Sanguineti, S. Toupin, S. Morisset, S. Champagne, J. Garot
- Change in left atrioventricular coupling index to predict hard cardiovascular disease: The Multi-Ethnic Study of Atherosclerosis (MESA)
- T. Pezel, B.A. Venkatesh, S.H. Heckbert, K. Yoko, H.D. De Vasconcellos, C.O. Wu, W. Post, D. Bluemke, D. Logeart, P. Henry, J. Lima
Bienvenue sur EM-consulte, la référence des professionnels de santé.