Computational Fluid Dynamics Study of the Inspiratory Upper Airway and Clinical Severity of Obstructive Sleep Apnea

Abstract

The apnea-hypopnea index (AHI) is a widely accepted measure for the severity of obstructive sleep apnea (OSA). Current methods to determine AHI fail to provide anatomic information for treatment decisions. In this report, we studied three-dimensional models of upper airways acquired by computed tomographic scanning with geometric measurements and computational fluid dynamics (CFD) analysis and evaluated the correlations with AHI.Participants had CT scans of their upper airways after standard polysomnography studies. Three-dimensional surface models of upper airways were generated for cross-sectional area measurements of the choanae (ACH) and the smallest cross-sectional area (Amin). Computational fluid dynamic analysis was then performed by using this three-dimensional model. Pressure differences required to set tidal volume during inspiration (ΔPmin-INSP) and expiration (ΔPmax-EXP) and minimum negative pressure produced in the level of ACH (Pmin-INSP at ACH) and Amin (Pmin-INSP at Amin) were calculated. Correlations of these parameters and the body mass index with AHI were analyzed. Statistical differences between groups of different AHI ranges were also compared.The pressure distribution simulated by CFD demonstrated abrupt pressure drops in Amin level, and this phenomenon was more significant in severe OSA. All parameters except ACH and Pmin-INSP at Amin significantly correlated with the AHI, and there were significant statistical differences between the OSA groups and the normal group. The results indicate that, in our study group, the geometry of pharyngeal airway and its CFD simulation correlate well with AHI. This model may be further applied for clinical evaluation.

Computational analysis of airflow dynamics for predicting collapsible sites in the upper airways: a preliminary study

“…Several studies have reported the use of CFD to simulate the process of airway obstruction in OSA (21,28,30,33,34). Most studies have exclusively used a fixed airflow volume and/or velocity for CFD.…”

Numerical study of dynamic glottis and tidal breathing on respiratory sounds in a human upper airway model

“…1). Similarly, Yu et al [30] predicted sharp pressure drops in the flow-limiting regions that were consistent with the apnea-hypopnea index (AHI). Large eddy simulations (LESs) of laryngeal flows included Lin et al [32] who observed significant effect of turbulent laryngeal jet on tracheal wall shear stress, Xi et al [33] who emphasized the importance of accurate glottis geometry in predicting airflows in the conducting airways, and Mihaescu et al [25] who demonstrated the superiority of large eddy simulation over Reynolds-averaged Navier-Stokes (RANS) methods using an image-based airway geometry representing severe OSA.…”

Changes in difficult airway predictors following mandibular setback surgery

“…17,18 The effective and reliable prediction of a difficult intubation requires a combination of several parameters, including age, BMI, Mallampati score, head and neck movement, thyromental and sternomental distances, dentition status, and inter-incisal distance. 19 In the current study, parameters among these that are considered practical, easily repeatable, and non-invasive were used.…”