Automated cardiac vortex ring identification and characterization based on Recurrent All-Pairs Field Transforms and...

Area of Science:

• Cardiovascular Imaging and Hemodynamics
• Computational Fluid Dynamics
• Medical Image Analysis

Background:

• Automated cardiac vortex identification is challenging due to complex intracardiac blood flow.
• Accurate characterization of blood flow patterns is crucial for understanding cardiac function.

Purpose of the Study:

• To develop and validate a novel algorithmic approach for automated identification and characterization of cardiac vortices.
• To improve the understanding of intracardiac hemodynamics and its impact on cardiac performance.

Main Methods:

• Utilized Recurrent All-Pairs Field Transforms (RAFT) to compute optical flow from Phase Contrast Magnetic Resonance Imaging (PC-MRI).
• Employed Lagrangian Averaged Vorticity Deviation (LAVD) for algorithmic characterization of vortex identification criteria.
• Applied Generalized Hough Transform (GHT) for automated structural depiction of cardiac vortices.

Main Results:

• The proposed method effectively identifies and describes the evolution of cardiac vortices.
• The RAFT framework accurately synthesizes intermediate PC-MRI data, reducing integral process errors in LAVD computation.
• Validation through cardiac flow data demonstrates the method's effectiveness and accuracy.

Conclusions:

• The developed method provides an effective means for automated cardiac vortex identification and analysis.
• Enhanced visualization and understanding of intracardiac hemodynamics are achievable for cardiologists.
• This approach holds potential for deeper insights into cardiac functional performance and related pathologies.