Understanding post-surgical decline in left ventricular function in primary mitral regurgitation using regression and machine learning models
Class I echocardiographic guidelines in primary mitral regurgitation (PMR) risks left ventricular ejection fraction (LVEF) < 50% after mitral valve surgery even with pre-surgical LVEF > 60%. There are no models predicting LVEF < 50% after surgery in the complex interplay of increased preloa...
Saved in:
Published in | Frontiers in cardiovascular medicine Vol. 10; p. 1112797 |
---|---|
Main Authors | , , , , , , , |
Format | Journal Article |
Language | English |
Published |
Switzerland
Frontiers Media S.A
21.04.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Class I echocardiographic guidelines in primary mitral regurgitation (PMR) risks left ventricular ejection fraction (LVEF) < 50% after mitral valve surgery even with pre-surgical LVEF > 60%. There are no models predicting LVEF < 50% after surgery in the complex interplay of increased preload and facilitated ejection in PMR using cardiac magnetic resonance (CMR).
Use regression and machine learning models to identify a combination of CMR LV remodeling and function parameters that predict LVEF < 50% after mitral valve surgery.
CMR with tissue tagging was performed in 51 pre-surgery PMR patients (median CMR LVEF 64%), 49 asymptomatic (median CMR LVEF 63%), and age-matched controls (median CMR LVEF 64%). To predict post-surgery LVEF < 50%, least absolute shrinkage and selection operator (LASSO), random forest (RF), extreme gradient boosting (XGBoost), and support vector machine (SVM) were developed and validated in pre-surgery PMR patients. Recursive feature elimination and LASSO reduced the number of features and model complexity. Data was split and tested 100 times and models were evaluated
stratified cross validation to avoid overfitting. The final RF model was tested in asymptomatic PMR patients to predict post-surgical LVEF < 50% if they had gone to mitral valve surgery.
Thirteen pre-surgery PMR had LVEF < 50% after mitral valve surgery. In addition to LVEF (
= 0.005) and LVESD (
= 0.13), LV sphericity index (
= 0.047) and LV mid systolic circumferential strain rate (
= 0.024) were predictors of post-surgery LVEF < 50%. Using these four parameters, logistic regression achieved 77.92% classification accuracy while RF improved the accuracy to 86.17%. This final RF model was applied to asymptomatic PMR and predicted 14 (28.57%) out of 49 would have post-surgery LVEF < 50% if they had mitral valve surgery.
These preliminary findings call for a longitudinal study to determine whether LV sphericity index and circumferential strain rate, or other combination of parameters, accurately predict post-surgical LVEF in PMR. |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 Reviewed by: Jiandong ZHOU, University of Oxford, United Kingdom Michal Jasinski, Wrocław University of Science and Technology, Poland Bharath Ambale Venkatesh, Johns Hopkins University, United States Edited by: Matteo Cameli, University of Siena, Italy Specialty Section: This article was submitted to Heart Valve Disease, a section of the journal Frontiers in Cardiovascular Medicine ORCID Jingyi Zheng orcid.org/0000-0002-0393-0997 Yuexin Li orcid.org/0000-0001-7425-8504 Nedret Billor orcid.org/0000-0003-4214-163X Mustafa I. Ahmed orcid.org/0000-0003-0885-8645 Yu-Hua Dean Fang orcid.org/0000-0002-1039-7983 Betty Pat orcid.org/0000-0003-1280-1457 Thomas S. Denney orcid.org/0000-0002-6695-4777 Louis J. Dell’Italia orcid.org/0000-0001-6654-0665 Abbreviations AUROC, area under the Receiver Operating Characteristic (ROC) curve; AUPRC, area under Precision-Recall (PR) curve; CMR, cardiac magnetic resonance; LV, left ventricle; LA, left atrium/atrial; LASSO, least absolute shrinkage and selection operator; LA EF, left atrial emptying fraction; LVEDD, left ventricular end-diastolic dimension; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic dimension; PMR, primary mitral regurgitation; RF, random forest; ROC, receiver operating characteristic; SHAP, Shapley Additive exPlanations; SVM, support vector machine; XGBoost. extreme gradient boosting; XO, xanthine oxidase. |
ISSN: | 2297-055X 2297-055X |
DOI: | 10.3389/fcvm.2023.1112797 |