DENSE-SIM

DENSE-SIM generates realistic cine DENSE images with high-resolution known ground-truth strain for evaluating accuracy and precision in strain analysis pipelines.

Key Features:

• Anatomy Generation: Creates end-diastolic cardiac shapes.
• Motion Generation: Produces myocardial deformations over time and Lagrangian strain.
• DENSE Image Generation: Uses Bloch equation simulations with realistic noise, spiral sampling, and phase cycling.
• Strain Evaluation: Enables comparison of estimated strain against ground-truth pixel strain.
• Modular Architecture: Comprises four main modules that can be used independently or together.

Scientific Applications:

• Strain Analysis Pipeline Validation: Evaluates accuracy and precision of strain estimation tools such as DENSEanalysis.
• Regularization Parameter Optimization: Assesses the impact of spatial regularization parameters on bias and variance characteristics in strain estimation.
• Deep Learning Development: Facilitates development of deep-learning models for myocardial strain analysis by providing realistic training data with known ground truth.
• Cardiac MRI Workflow Enhancement: Supports clinical cardiovascular magnetic resonance workflows through improved strain analysis tool evaluation.

Methodology:

The pipeline comprises four modules: anatomy generation for end-diastolic cardiac shapes, motion generation for myocardial deformations and Lagrangian strain, DENSE image generation using Bloch equation simulations with realistic noise, spiral sampling, and phase cycling, and strain evaluation against ground-truth data.