Dynamics of the human upper airway: On the development of a three-dimensional computational model

The advances reported herein form part of a larger project that has as its objective the development of a full flow-structure-interaction model of the human upper airway. Here we first briefly report on a two-dimensional (saggital section) model built using published CT-scan geometric data. For the development of our three-dimensional capability, we use the unique data captured in vivo by an endoscopic optical technique that we have developed. This measurement system, described as anatomical optical coherence tomography (aOCT), allows quantitative real-time imaging of the internal anatomy of the human upper airway with minimal invasiveness. Moreover, the system permits motions of the internal geometry at a fixed location to be recorded. The aOCT data set is insufficient by itself to construct a complete geometry because only the polar coordinates are obtained in a local reference frame. Accordingly, the locus described by the endoscope, in which the aOCT is housed, is obtained by orthogonal CT scans. The combination of CT scans and aOCT measurements then provides the required geometric information for the construction of the computational model. Results of a twodimensional model show how the soft palate responds to the mean-flow variations of the breathing cycle. For the threedimensional work, the key results of this paper rest in the reconstruction of the time-dependent geometry of the upper airway, the first time that this has been accomplished using direct internally-based measurement