3D phase contrast MRI in models of human airways: Validation of computational fluid dynamics simulations of steady inspiratory flow

Abstract

Background: Knowledge of airflow patterns in the large airways is of interest in obstructive airways disease and in the development of inhaled therapies. Computational fluid dynamics (CFD) simulations are used to study airflow in realistic airway models but usually need experimental validation.

Purpose: To develop MRI-based methods to study airway flow in realistic 3D-printed models.

Study type: Case control.

Phantom: Two 3D-printed lung models.

Field strength/sequence: 1.5-3T, flow MRI.

Assessment: Two human airway models, respectively including and excluding the oral cavity and upper airways derived from MR and CT imaging, were 3D-printed. 3D flow MRI was performed at different flow conditions corresponding to slow and steady airflow inhalation rates. Water was used as the working fluid to mimic airflow. Dynamic acquisition of 1D velocity profiles was also performed at different locations in the trachea to observe variability during nonsteady conditions.

Statistical tests: Linear regression analysis to compare both flow velocity fields and local flow rates from CFD simulations and experimental measurement with flow MRI.

Results: A good agreement was obtained between 3D velocity maps measured with flow MRI and predicted by CFD simulations, with linear regression R-squared values ranging from 0.39 to 0.94 when performing a pixel-by-pixel comparison of each velocity component. The flow distribution inside the lung models was also similar, with average slope and R-squared values of 0.96 and 0.99, respectively, when comparing local flow rates assessed at different branching locations. In the model including the upper airways, a turbulent laryngeal jet flow was observed with both methods and affected remarkably the velocity profiles in the trachea.

Data conclusion: We propose flow MRI using water as a surrogate fluid to air, as a validation tool for CFD simulations of airflow in geometrically realistic models of the human airways.

Level of evidence: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:1400-1409.

Keywords: 3D-printing model; CFD; airflow; lungs.