3D harmonic phase tracking with anatomical regularization.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Cardiovascular Research

Background:

• Accurate assessment of cardiac mechanics is crucial for diagnosing cardiovascular diseases.
• Traditional methods for analyzing 3D tagged MRI data have limitations in precision and computational efficiency.
• Quantifying myocardial strain, particularly radial strain, remains a challenge in cardiac imaging.

Purpose of the Study:

• To develop and validate a novel algorithm extending the Hierarchical Analysis of Regional Motion (HARP) for 3D tagged MRI.
• To improve the accuracy of tracking left ventricular (LV) displacements and calculating myocardial strain.
• To integrate anatomical regularization and myocardium incompressibility for enhanced strain recovery.

Main Methods:

• Developed a novel regularization framework in an anatomical coordinate space for HARP.
• Incorporated myocardium incompressibility to correct radial strain measurements.
• Utilized a volumetric mesh for computationally efficient tracking and regularization of LV displacements.
• Extended a window-weighted regression method for robust cardiac motion tracking at various scales.

Main Results:

• Achieved tracking accuracy comparable to top methods in a recent benchmark on healthy volunteers.
• Demonstrated low bias and strain errors (<5% for longitudinal/circumferential, (-5%,5%) for radial) on synthetic data.
• Observed correlation between strain dispersion and transmural fibrosis extent in clinical data.
• Identified reduced deformation values within infarcted myocardial segments.

Conclusions:

• The novel algorithm provides accurate and robust quantification of 3D cardiac motion and strain from tagged MRI.
• The method shows potential for improved diagnosis and characterization of myocardial pathologies, including fibrosis and infarction.
• Integration of anatomical regularization and physical constraints enhances the reliability of strain measurements.