Enhanced Extraction of Activation Time and Contractility From Myocardial Strain Data Using Parameter Space Features...

Area of Science:

• Computational Biology
• Biomedical Engineering
• Cardiovascular Physiology

Background:

• Accurate extraction of myocardial activation time (AT) and contractility (Con) from cardiac strain is crucial for understanding heart function.
• Current computational models face challenges with parameter interference, reducing extraction precision and accuracy.

Purpose of the Study:

• To investigate if leveraging parameter space features can enhance the accuracy of AT and Con extraction from cardiac strain.
• To develop a method that accounts for the asymmetric distribution of parameter values in the parameter space.

Main Methods:

• A computational model simulating sarcomere mechanics was used to create a parameter space grid of AT and Con pairs.
• Synthetic fingerprints (binary images) were generated from simulated strain patterns to represent extraction uniqueness.
• A measurement fingerprint from patient data was compared to synthetic fingerprints to create a proximity map for enhanced parameter extraction.

Main Results:

• The enhanced parameter extraction method using a proximity map significantly improved accuracy compared to simple optimization.
• Extracted AT values improved from -59 to 19 ms to -16 to 14 ms.
• Extracted Con values improved from 48% to 110% to 85% to 110%.

Conclusions:

• Leveraging parameter space features, specifically through proximity maps, effectively addresses the asymmetric distribution of parameter values.
• This approach significantly enhances the accuracy and precision of extracting myocardial activation time and contractility from cardiac strain data.