Patient specific modeling will be conducted in two parts. To form a basis for the research, a 3D CFD model of the upper airways will first be established. The geometry of the model will be based on CT/MRI data obtained from WP1 and rigid walls will be assumed. By performing CFD simulations of specific patients, it will be possible to correlate key flow parameters to the patient response to treatment. The second part consists of implementing a FSI model of the upper airways to be able to assess the impact of the two-way coupling between the fluid and tissue stresses and the deformable fluid-tissue interface.
When an adequate CFD model has been obtained for a specific situation, the CFD model is a powerful tool for predicting the effect of surgical modification of the geometry of the nasal cavity. CFD modeling consists of three main ingredients; 1) geometry modeling, 2) fluid dynamics, and 3) validation of the model. WP4 will aim at fulfilling these three outlined tasks, with the ultimate goal of providing a platform for technologically based OSAS diagnostics and mitigation. It is realized, however, that a full FSI model is required to get an accurate description of the air flow. But currently no FSI method solves this problem with high speed or accuracy. The major contribution to the time-consumption of FSI modeling is the remapping of the computational mesh. Thus, a major advantage would be achieved by being able to utilize fixed grid methods such as the immersed boundary method. WP4 will investigate the possibilities of employing results from WP3 in developing sub-grid models for the boundary conditions and the momentum exchange that takes place at the wall, to enable the use of fixed grid methods. This will reduce the number of simplifying assumptions required to obtain a converged solution.
The mathematical modeling of biomechanical systems is characterized by the complexity of the computational domain, the complex dynamics that arises in the two-way interactions between the flowing air and the solid but plastic/elastic biological tissue, and the multi-scale (length and time) nature of biological transport mechanisms. The commercial CFD software Ansys Fluent will be utilized to solve for the flow field in the upper airways. Fluent also has FSI capabilities built in. Sub-grid models to handle the two-way coupling between fluid and tissue, will be implemented as user-defined functions (UDF) in Fluent. The developed subgrid models will then be benchmarked against Ansys Fluents built-in FSI model. Significant numerical challenges need to be overcome due to the complex geometry, transient boundary conditions and airflow, and two-way coupling between dynamic air and tissue stresses and solid boundary deformations.
In bio-mechanical applications, obtaining models for the geometry of the computational domain is feasible by the combination of CT, for the bony structures, and MRI, for the soft tissue. For converting multi-slice 2D CT/MRI data into 3D CAD format, academically licensed software will be utilized by master students and/or the PhD student at NTNU. Relevant software includes Segment, Mimics and Simpleware. Based on data obtained in WP1, patient specific geometries will be generated before and after nasal surgery to form a basis for CFD analysis of the qualitative result of surgery and quantitative effect on flow parameters.
For validation of the developed models, a simplified approach is taken. Experiments and flow measurements will be performed in a simplified physical model of the upper airways, including the mouth and nasal cavities, pharynx and a soft palate made from a flexible material of known elastic properties. The physical model will be built in a transparent material so that the flow-induced vibrations in the soft palate can be analyzed by video recordings. The simplified model will act as a calibration standard for models developed in both WP3 and WP4. The experiments will be carried out as part of a master of technology project at NTNU.
Based on the 3D CAD models of the upper airways, it is possible to utilize NTNU/SINTEF expertise on 3D printing to produce physical models of e.g. the pharyngeal tract. Flow measurements may then be performed in the physical model to calibrate/validate the CFD models with rigid walls.