Particle Tracking through the Human Respiratory System

Notable Features of Model

• Based on in vivo patient data
• CFD analysis

Purpose of the Study

The study of airflow through the human respiratory system is a complex problem, which has in the past necessitated the use of numerical models based on grossly simplified geometry. The aims of the study were (a) to compare an idealized geometry with that of a physiologically realistic geometry and (b) to examine the flow of air and track the trajectories of water vapour particles, through the human respiratory system.

Mesh in +ScanFE

CFD idealized geometry

CFD from MRI scan

Scanning and segmentation in ScanIP

In vivo MRI scans were performed at a resolution of 1.44 x 1.44 x 1.44 mm.

ScanIP was used to create a physiologically realistic model of the human respiratory system from the MRI data. In order to create a 3D representation of just the airways, first they had to be defined against their surroundings. In ScanIP a combination of Threshold and FloodFill tools were used to create a mask of the airways based on the grey scale values of the image.

Mesh Generation in ⁺ScanFE

In ⁺ScanFE the first operation was to define the contact surfaces of each part that was intended for export. In the case of the human airways, three contact surface needed to be created:
(1) contact from the airways to the background which would become the walls of the airways,
(2) contact from the airways to the exterior which would become the inlet and outlet of the airways, and
(3) contact between the airways and themselves.

Once the contact surfaces were defined, the mesh was created within ⁺ScanFE and exported in Fluent format.

CFD Analysis in Fluent

In Fluent, calculations were performed using a turbulent flow rate of 0.0012 m³ s^-1 under both aspiratory and expiratory flow conditions. There was very little agreement between the flow structures of the idealised geometry and that obtained from the MRI scan data. Re-circulatory regions in the idealised geometry were induced by sharp edges in the geometry but none of these re-circulatory flows were found to exist in the MRI geometry. It was evident that the curved features of the MRI geometry helped keep the flow attached to the walls thus preventing separation, re-circulation and adverse pressure gradients.

A tracking study of a range of water vapour particles with a median diameter of 5.0 µm and a 90th percentile diameter of 10.0 µm, showed that small variations in particle diameter had little effect over the eventual destinations of the particles whereas the zones of the respiratory system with the most complex and curving geometry, trapped a higher percentage of larger diameter particles. These findings agree with those found previously in relation to vapour distribution from medical nebulizers.