Message Passing Interface Parallelization for Two-Way Coupled Euler-Lagrange Simulation of Microbubble Enhanced HIFU

Area of Science:

• Computational physics and biomedical engineering.
• Advanced modeling of ultrasound therapies.

Background:

• Microbubble-enhanced high-intensity focused ultrasound (HIFU) shows promise for liver and brain cancer tumor ablation.
• Accurate characterization of acoustic and thermal fields is crucial for optimizing HIFU treatments.

Purpose of the Study:

• To develop and validate a parallelized computational model for microbubble-enhanced HIFU.
• To accurately simulate the coupled acoustic, thermal, and microbubble dynamics during HIFU procedures.

Main Methods:

• A coupled Euler-Lagrange model was employed, solving Navier-Stokes equations for the fluid and tracking microbubbles in a Lagrangian manner.
• Message Passing Interface (MPI) parallelization using domain decomposition was implemented for efficient computation.
• Ghost cells were utilized to manage the coupling of bubble effects across subdomain borders, minimizing communication overhead.

Main Results:

• The proposed parallelization scheme was verified for gas effects conservation and validated on a microbubble-enhanced HIFU scenario.
• Demonstrated efficient parallelization scaling and performance analysis for the computational model.
• The ghost cell strategy effectively handled inter-subdomain coupling without direct bubble information exchange.

Conclusions:

• The developed parallelized Euler-Lagrange model provides an efficient and accurate tool for simulating microbubble-enhanced HIFU.
• This computational approach can aid in the precise characterization and optimization of HIFU for cancer treatments.