UBC Theses and Dissertations
A novel SPH method for investigating the role of saliva in swallowing using 4D CT images Ho, Andrew Kenneth
The thesis presents novel computer methods towards simulation of oropha- ryngeal swallowing. The anatomy and motion of the human upper airway was extracted from dynamic Computed Tomography (CT) data using a novel tool and workflow. A state-of-the-art SPH method is extended to accommo- date non-Newtonian materials in the extracted geometries. A preliminary numerical experiment of six human oropharyngeal swallows using Smoothed Particle Hydrodynamics (SPH) demonstrates that the methods are robust and useful for simulation of oropharyngeal swallowing. The presence of saliva is well known to be important for mastication, swallowing, and overall oral health. However, clinical studies of patients with hyposalivation are unable to isolate the effect of saliva from other con- founding factors. The simulation presented in this thesis examines fluid boluses under lubricated and non-lubricated boundary conditions. Upon comparison with medical image data, the experiments suggest that saliva does not provide a significant lubricative effect on the bolus transit times, but it may serve to reduce residue and therefore improve overall swallowing efficacy. Our findings, while preliminary, corroborate with existing clinical research that finds that groups with hyposalivation do not have significantly different transit times with control groups, but that residue may be increased in the hyposalivation group. Previous studies using computer simulation of fluid flow in the orophar- ynx typically make use of simplified geometries. Our work uses dynamic 320-row Area Detector Computed Tomography (ADCT) images as the ba- sis for the simulations, and therefore does not require simplifying geometric assumptions. Since the data are dynamic, motion trajectories are all sup- plied by the ADCT data, and extrapolation from 2D sources such as bi-plane videofluoroscopy is not required. Processing the image data required the de- velopment of a novel workflow based on a new tool, which we call BlendSeg. We utilize and extend Unified Semi-Analytic Wall (USAW) SPH methods so that orophrayngeal swallowing simulations may be performed. These extensions include the simulation of non-Newtonian boluses, and moving 3D boundaries. Partial validation of the extended USAW SPH method is performed using canonical flows.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International