Joint statistics on cardiac shape and fiber architecture

Area of Science:

• Cardiovascular Science
• Biomedical Engineering
• Medical Imaging

Background:

• Cardiac fiber architecture is crucial for heart function but its variability and relation to myocardial shape are not fully understood.
• Current methods often rely on fixed average geometries, limiting the analysis of inter-subject differences.

Purpose of the Study:

• To extend statistical analysis of cardiac fiber architecture beyond fixed average geometries.
• To investigate the co-variation between cardiac fiber architecture and myocardial shape or strain.
• To improve the synthesis of cardiac fiber architecture from patient-specific geometry.

Main Methods:

• Developed a statistical framework to analyze joint variations between cardiac fiber architecture and shape/strain data.
• Applied the framework to a dataset of 8 ex vivo canine hearts.
• Compared joint statistics approach with traditional atlas-based methods.

Main Results:

• Strain-based information demonstrated the strongest correlation with cardiac fiber architecture.
• The joint statistics approach improved fiber synthesis from shape by 8.0% compared to current methods, with some cases showing up to 25.9% improvement.
• Evidence suggests a link between architectural variability and myocardial thickness.

Conclusions:

• Joint statistical analysis offers a more accurate method for understanding and synthesizing cardiac fiber architecture.
• Strain-based data is a key factor in predicting cardiac fiber architecture.
• Further research into the relationship between architectural variability and myocardial thickness is warranted.