Machine Learning Outcome Prediction in Dilated Cardiomyopathy Using Regional Left Ventricular Multiparametric Strain
- PDF / 1,242,222 Bytes
- 11 Pages / 593.972 x 792 pts Page_size
- 84 Downloads / 201 Views
Annals of Biomedical Engineering ( 2020) https://doi.org/10.1007/s10439-020-02639-1
Original Article
Machine Learning Outcome Prediction in Dilated Cardiomyopathy Using Regional Left Ventricular Multiparametric Strain ROBERT M. MACGREGOR ,1 AIXIA GUO,2 MUHAMMAD F. MASOOD,1 BRIAN P. CUPPS,1 GREGORY A. EWALD,3 MICHAEL K. PASQUE,1 and RANDI FORAKER2 1 Department of Surgery, Division of Cardiothoracic Surgery, Barnes-Jewish Hospital, Washington University School of Medicine, Campus Box 8234, 660 S. Euclid Ave., St. Louis, MO 63110, USA; 2Institute for Informatics, Division of General Medical Sciences, Washington University School of Medicine, St. Louis, MO, USA; and 3John T. Milliken Department of Internal Medicine, Cardiovascular Division, Washington University School of Medicine, St. Louis, MO, USA
(Received 9 June 2020; accepted 24 September 2020) Associate Editor Lakshmi Prasad Dasi oversaw the review of this article.
Abstract—The clinical presentation of idiopathic dilated cardiomyopathy (IDCM) heart failure (HF) patients who will respond to medical therapy (responders) and those who will not (non-responders) is often similar. A machine learning (ML)-based clinical tool to identify responders would prevent unnecessary surgery, while targeting non-responders for early intervention. We used regional left ventricular (LV) contractile injury patterns in ML models to identify IDCM HF nonresponders. MRI-based multiparametric strain analysis was performed in 178 test subjects (140 normal subjects and 38 IDCM patients), calculating longitudinal, circumferential, and radial strain over 18 LV sub-regions for inclusion in ML analyses. Patients were identified as responders based upon symptomatic and contractile improvement on medical therapy. We tested the predictive accuracy of support vector machines (SVM), logistic regression (LR), random forest (RF), and deep neural networks (DNN). The DNN model outperformed other models, predicting response to medical therapy with an area under the receiver operating characteristic curve (AUC) of 0.94. The top features were longitudinal strain in (1) basal: anterior, posterolateral and (2) mid: posterior, anterolateral, and anteroseptal sub-regions. Regional contractile injury patterns predict response to medical therapy in IDCM HF patients, and have potential application in ML-based HF patient care.
Keywords—Heart failure, Regional contractile injury, Magnetic resonance imaging, Machine learning, Deep learning, Myocardial strain.
NOMENCLATURE AUC DENSE DNN EHR HF IDCM LR LV ML NYHA RF ROC SVM
Area under the curve Displacement encoding with stimulated echoes Deep neural networks Electronic health record Heart failure Idiopathic dilated cardiomyopathy Logistic regression Left ventricle Machine learning New York Heart Association Random forest Receiver operating characteristic curve Support vector machines
INTRODUCTION Address correspondence to Michael K. Pasque, Department of Surgery, Division of Cardiothoracic Surgery, Barnes-Jewish Hospital, Washington University School of Medi
Data Loading...
