Author + information
- Received January 22, 2018
- Revision received April 17, 2018
- Accepted April 26, 2018
- Published online June 13, 2018.
- Manar D. Samad, PhDa,
- Alvaro Ulloa, MSa,
- Gregory J. Wehner, PhDb,
- Linyuan Jing, PhDa,
- Dustin Hartzel, BSa,
- Christopher W. Good, DOc,
- Brent A. Williams, PhDd,
- Christopher M. Haggerty, PhDa and
- Brandon K. Fornwalt, MD, PhDa,b,e,∗ ()
- aDepartment of Imaging Science and Innovation, Geisinger, Danville, Pennsylvania
- bDepartment of Biomedical Engineering, University of Kentucky, Lexington, Kentucky
- cDepartment of Cardiology, Geisinger, Danville, Pennsylvania
- dDepartment of Epidemiology and Health Services Research, Geisinger, Danville, Pennsylvania
- eDepartment of Radiology, Geisinger, Danville, Pennsylvania
- ↵∗Address for correspondence:
Dr. Brandon K. Fornwalt, Geisinger, 100 North Academy Avenue, Danville, Pennsylvania, 17822-4400.
Objectives The goal of this study was to use machine learning to more accurately predict survival after echocardiography.
Background Predicting patient outcomes (e.g., survival) following echocardiography is primarily based on ejection fraction (EF) and comorbidities. However, there may be significant predictive information within additional echocardiography-derived measurements combined with clinical electronic health record data.
Methods Mortality was studied in 171,510 unselected patients who underwent 331,317 echocardiograms in a large regional health system. We investigated the predictive performance of nonlinear machine learning models compared with that of linear logistic regression models using 3 different inputs: 1) clinical variables, including 90 cardiovascular-relevant International Classification of Diseases, Tenth Revision, codes, and age, sex, height, weight, heart rate, blood pressures, low-density lipoprotein, high-density lipoprotein, and smoking; 2) clinical variables plus physician-reported EF; and 3) clinical variables and EF, plus 57 additional echocardiographic measurements. Missing data were imputed with a multivariate imputation by using a chained equations algorithm (MICE). We compared models versus each other and baseline clinical scoring systems by using a mean area under the curve (AUC) over 10 cross-validation folds and across 10 survival durations (6 to 60 months).
Results Machine learning models achieved significantly higher prediction accuracy (all AUC >0.82) over common clinical risk scores (AUC = 0.61 to 0.79), with the nonlinear random forest models outperforming logistic regression (p < 0.01). The random forest model including all echocardiographic measurements yielded the highest prediction accuracy (p < 0.01 across all models and survival durations). Only 10 variables were needed to achieve 96% of the maximum prediction accuracy, with 6 of these variables being derived from echocardiography. Tricuspid regurgitation velocity was more predictive of survival than LVEF. In a subset of studies with complete data for the top 10 variables, multivariate imputation by chained equations yielded slightly reduced predictive accuracies (difference in AUC of 0.003) compared with the original data.
Conclusions Machine learning can fully utilize large combinations of disparate input variables to predict survival after echocardiography with superior accuracy.
This work was supported by a National Institutes of Health Director’s Early Independence Award (DP5 OD-012132). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Received January 22, 2018.
- Revision received April 17, 2018.
- Accepted April 26, 2018.
- 2018 American College of Cardiology Foundation
This article requires a subscription or purchase to view the full text. If you are a subscriber or member, click Login or the Subscribe link (top menu above) to access this article.