Generating Purkinje networks in the human heart

Area of Science:

• Cardiovascular Research
• Computational Biology
• Medical Imaging

Background:

• The Purkinje network is crucial for cardiac excitation, but in vivo imaging is lacking.
• Computational models offer an alternative for visualizing and understanding the Purkinje system.
• Existing models struggle with manual generation and irregular ventricular surfaces.

Purpose of the Study:

• To develop a novel algorithm for robust Purkinje network reconstruction in the human heart.
• To provide a freely available computational tool for cardiac modeling.
• To demonstrate the impact of a dense Purkinje network on cardiac electrophysiology simulations.

Main Methods:

• Developed a new algorithm integrating fractal tree generation with a projection method for irregular surfaces.
• Utilized Monte Carlo simulations to generate and analyze Purkinje network density and structure.
• Performed cardiac electrophysiology simulations comparing the new network, a left-sided network, and no network.

Main Results:

• The new algorithm generated significantly denser Purkinje networks (1219±61 branches) compared to conventional methods (419±107 branches).
• Achieved twice the surface density (11±3mm coverage) compared to conventional approaches (22±7mm).
• Simulations with the new network predicted more realistic activation sequences and times, aligning with clinical ECGs in various conditions.

Conclusions:

• The developed algorithm reliably creates robust and dense Purkinje networks.
• Dense Purkinje networks are essential for accurate human heart electrophysiology simulations.
• This work supports personalized cardiac interventions like pacing and ablation for rhythm restoration.