Subject-specific modeling framework for particle deposition using computational fluid dynamics

Area of Science:

• Computational modeling and simulation
• Respiratory physiology
• Medical imaging and image analysis

Background:

• Accurate quantification of particle deposition and dose in the respiratory tract is crucial for health and safety.
• Existing computational models lack detailed deposition profiles and subject-specific capabilities.
• Manual segmentation and preprocessing for simulations are time-consuming and limit accessibility.

Purpose of the Study:

• To develop a fully automated workflow for individualized particle deposition profiles in the human respiratory tract.
• To integrate advanced computer vision and computational fluid dynamics for high-fidelity simulations.
• To streamline the process from CT imaging to dose assessment.

Main Methods:

• Automated segmentation of respiratory tract geometries from CT scans using deep learning.
• Preprocessing algorithms for geometry quality checking, artifact correction, and mesh generation.
• Computational Fluid and Particle Dynamics (CFPD) simulations using OpenFOAM or StarCCM+ under realistic breathing conditions.

Main Results:

• Successful generation of individualized 3D respiratory tract models.
• Automated preprocessing pipeline reducing manual intervention for CFPD simulations.
• High-fidelity particle deposition profiles and dose distributions calculated for personalized assessments.

Conclusions:

• The automated workflow significantly improves access to subject-specific respiratory particle deposition modeling.
• Enhanced precision in particle deposition and dose calculations can inform personalized respiratory treatments.
• The framework refines dose estimates for radiation protection and aids in understanding aerosol behavior.