Cardiac Magnetic Resonance-Derived Parametric Mapping Radiomics for Prediction of Left Ventricular Reverse Remodeling...

Area of Science:

• Cardiology
• Radiology
• Medical Imaging

Background:

• Nonischemic dilated cardiomyopathy (DCM) is a significant cause of heart failure.
• Left ventricular reverse remodeling (LVRR) is a key indicator of improved prognosis in DCM patients.
• Predicting LVRR is crucial for guiding treatment strategies and patient management.

Purpose of the Study:

• To evaluate the predictive capability of radiomic scores (RadScores) derived from cardiac magnetic resonance (CMR) parametric mapping for LVRR in nonischemic DCM.
• To assess the incremental value of RadScores beyond conventional imaging markers.

Main Methods:

• Retrospective analysis of 74 nonischemic DCM patients undergoing CMR.
• Calculation of RadScores using logistic regression on native T1 and extracellular volume (ECV) fraction maps.
• Definition of LVRR based on changes in left ventricular ejection fraction post-CMR.
• Evaluation of predictive performance using receiver operating characteristic (ROC) analysis.

Main Results:

• Patients achieving LVRR showed significantly higher T1-RadScore and ECV-RadScore compared to non-responders.
• T1-RadScore achieved an area under the ROC curve (AUC) of 0.834 for LVRR prediction.
• Combining T1- and ECV-RadScores with late gadolinium enhancement significantly improved LVRR prediction (AUC 0.903).

Conclusions:

• Radiomic scores from native T1 and ECV CMR maps show promise in predicting LVRR in nonischemic DCM.
• RadScores provide incremental prognostic information beyond late gadolinium enhancement.
• These findings may lead to improved risk stratification and personalized treatment in DCM.