This presentation was made at CAASE18, The Conference on Advancing Analysis & Simulation in Engineering. CAASE18 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, to share experiences, discuss relevant trends, discover common themes, and explore future issues.

Resource Abstract

Purpose
Obstructive sleep apnea (OSA) is a serious medical condition that has considerable health and social consequences. OSA occurs as a person enters sleep and the muscles in the throat relax leading to localized airway collapse. Mandibular repositioning devices (MRD) are an effective, non-invasive treatment option for patients with OSA. The current study uses computational fluid dynamics (CFD) to predict the flow through airways of patients with OSA, both with and without a MRD. The goal being to identify flow and anatomy-related parameters that best correlate the CFD predictions with patient outcomes.

Methods
The severity of OSA is diagnosed using a polysomnography (PSG) test conducted and was analyzed at the San Francisco Veterans Affairs Medical Center (SFVA) Sleep Center. Based on the severity, patients were selected to have a custom MRD fabricated. For these patients, a repeat PSG with the MRD in place was obtained. These patients also underwent a CT scan with and without the MRD. Three-dimensional models of the airways were reconstructed from the CT scans using MIMICS software (Materialise, Ann Arbor, MI). Steady state CFD simulations using ANSYS-CFX (ANSYS Inc, Canonsburg, PA) at a near peak inhalation air flow rate of 300 ml/sec were used for this study. Airway compliance was not included in these simulations.

Results
For this pilot study, a total of eight patients were treated with a MRD. Based upon their PSG results, each patient showed improvement in OSA severity as measured by their change in apnea-hypopnea index. The CFD results focused on the predicting the flow through the soft palate and behind the epiglottis. These are the two airway locations that typically exhibit the smallest opening size and therefore greatest likelihood of airway collapse. Key parameters studied included the local flow areas, anterior & posterior clearances, peak airway velocity, overall airway resistance, and minimum static pressures along the back of the airways. For five of the patients, these key parameters all showed consistent improvement with the MRD. The key parameters trended in opposite directions for the other three patients.

Conclusions
The CFD results predicted improvement using a MRD for five of the eight patients studied. For the patients with predicted improvement, the increase in flow area near the soft palate was dramatic, particularly in the lateral dimension. Such information was made clearer when viewing the three-dimensional airway reconstructions. The CFD results provide quantitative information concerning the change in local flow velocities and static pressures along the airway. A lower static pressure and narrower airway suggesting an increased likelihood of airway collapse. An interesting observation, for the three patients where the CFD modeling and PSG testing disagreed, is the amount of mandible advancement was near the upper range for this group of patients. Our future focus is on expanding the patient database and looking more closely at the anatomical changes and variability in patient CT scans with use of a MRD.