LGE

LGE analyzes late gadolinium-enhanced (LGE) cardiac magnetic resonance images to quantify myocardial scar morphology and simulate electrophysiological behavior to investigate arrhythmogenic substrates in non-ischemic dilated cardiomyopathy.

Key Features:

• Automated scar extraction: Extracts morphological features from short-axis LGE MRI images provided in NIfTI format.
• Scar area: Computes total area covered by fibrotic scarring.
• Entropy: Quantifies heterogeneity and complexity of scar intensity patterns.
• Components: Identifies and counts distinct scar regions within each image.
• Transmurality: Measures percentage depth of scar penetration through the myocardial wall.
• Radiality: Quantifies angular extent of LGE around the blood-pool center on a scale from 0 to 1.
• Interface length: Calculates the total boundary length between healthy myocardium and scar tissue.
• Electrophysiological simulation: Generates two-dimensional electrophysiological models from LGE images and applies programmed electrical stimulation to assess propensity for transmural block or reentrant circuits.
• Microstructural assignment: Assigns LGE regions to one of ten microstructural scar models at sub-MRI resolution for analysis of scar morphology effects on electrophysiology.

Scientific Applications:

• Arrhythmia prediction: Correlates morphological features with simulation outcomes to predict reentrant circuits and transmural block.
• Risk stratification: Differentiates scar patterns that lead to electrical block versus those that promote reentry to inform patient risk assessment.
• Pathophysiological research: Investigates relationships between cardiac fibrosis microstructure and electrical instability in non-ischemic dilated cardiomyopathy.

Methodology:

Processes 699 short-axis LGE MRI images (NIfTI) from 157 patients, assigns LGE regions to one of ten microstructures, constructs two-dimensional electrophysiological models, applies simulated programmed electrical stimulation to assess transmural block or reentrant circuits, and analyzes outcomes alongside extracted morphological features using linear discriminant analysis to identify predictors of block and reentry.