Assessing Myocardial Microstructure With Biophysical Models of Diffusion MRI

Area of Science:

• Biophysics
• Medical Imaging
• Cardiovascular Science

Background:

• Biophysical models offer insights into tissue microstructure from diffusion weighted magnetic resonance imaging (DW-MRI).
• Cardiac microstructure mapping (CMM) has faced limitations in applying these advanced models.
• Existing models struggle to accurately represent complex cardiac tissue structures.

Purpose of the Study:

• To develop and evaluate novel two-compartment biophysical models for cardiac microstructure mapping.
• To represent intra- and extracellular spaces using combinations of restricted cylinder models and diffusion tensors.
• To assess the performance of extended cylinder models (ECS, GDR, BDA) in cardiac tissue.

Main Methods:

• Proposed seven new two-compartment models combining restricted cylinder models with diffusion tensors.
• Applied models to fixed mouse heart DW-MRI data with varying diffusion times, q-shells, and directions.
• Utilized high-resolution 3D synchrotron X-ray imaging (SRI) for model validation and developed a novel tensor-based registration method.

Main Results:

• The cylinderGDR-pancake model achieved a 25% reduction in root mean squared error (RMSE) compared to diffusion tensor imaging (DTI).
• The cylinderBDA-pancake model demonstrated superior anatomical accuracy and the ability to model dispersion.
• Consistent parameters between SRI and DW-MRI validated the biophysical models' efficacy.

Conclusions:

• Novel compartment models significantly improve cardiac microstructure mapping accuracy using DW-MRI.
• The cylinderGDR-pancake and cylinderBDA-pancake models show strong potential for detailed cardiac tissue analysis.
• Biophysical models, particularly compartment models, hold promise for assessing physiologically relevant cardiac microstructure parameters.