Reproducible Classification of Infarct Heterogeneity Using Fuzzy Clustering on Multicontrast Delayed Enhancement Magnetic Resonance Images

Abstract

Delayed enhancement MRI (DE-MRI) can be used to identify myocardial infarct (MI). Classification of MI into the infarct core and heterogeneous periphery (called the gray zone) on conventional inversion-recovery gradient echo (IR-GRE) DE-MRI images has been related to inducibility for ventricular tachycardia. However, this classification is sensitive to image noise, depends on the signal intensity characteristics in a remote region of myocardium, and requires manual contours of the endocardial border. Image analysis and fuzzy clustering techniques were developed to analyze images acquired using a multicontrast delayed enhancement (MCDE) sequence in order characterize the infarct zones. The MCDE analysis is automated and uses data fitting of signal intensities acquired at multiple inversion times. In a study of 15 patients with chronic MI, the gray zones derived from IR-GRE and MCDE images were comparable. The variability in the gray zone size associated with random noise and operator input was significantly reduced using the MCDE-based analysis compared to the IR-GRE-based analysis. In summary, the MCDE approach yields a more reproducible measure of the infarct core and gray zones on any given data set.