Login / Signup

Velopharyngeal mucosal surface topography in healthy subjects and subjects with obstructive sleep apnea.

Christopher LambethJason AmatouryZiyu WangSheryl FosterTerence AmisKristina Kairaitis
Published in: Journal of applied physiology (Bethesda, Md. : 1985) (2016)
Macroscopic pharyngeal anatomical abnormalities are thought to contribute to the pathogenesis of upper airway (UA) obstruction in obstructive sleep apnea (OSA). Microscopic changes in the UA mucosal lining of OSA subjects are reported; however, the impact of these changes on UA mucosal surface topography is unknown. This study aimed to 1) develop methodology to measure UA mucosal surface topography, and 2) compare findings from healthy and OSA subjects. Ten healthy and eleven OSA subjects were studied. Awake, gated (end expiration), head and neck position controlled magnetic resonance images (MRIs) of the velopharynx (VP) were obtained. VP mucosal surfaces were segmented from axial images, and three-dimensional VP mucosal surface models were constructed. Curvature analysis of the models was used to study the VP mucosal surface topography. Principal, mean, and Gaussian curvatures were used to define surface shape composition and surface roughness of the VP mucosal surface models. Significant differences were found in the surface shape composition, with more saddle/spherical and less flat/cylindrical shapes in OSA than healthy VP mucosal surface models (P < 0.01). OSA VP mucosal surface models were also found to have more mucosal surface roughness (P < 0.0001) than healthy VP mucosal surface models. Our novel methodology was utilized to model the VP mucosal surface of OSA and healthy subjects. OSA subjects were found to have different VP mucosal surface topography, composed of increased irregular shapes and increased roughness. We speculate increased irregularity in VP mucosal surface may increase pharyngeal collapsibility as a consequence of friction-related pressure loss.NEW & NOTEWORTHY A new methodology was used to model the upper airway mucosal surface topography from magnetic resonance images of patients with obstructive sleep apnea and healthy adults. Curvature analysis was used to analyze the topography of the models, and a new metric was derived to describe the mucosal surface roughness. Increased roughness was found in the obstructive sleep apnea vs. healthy group, but further research is required to determine the functional effects of the measured difference on upper airway airflow mechanics.
Keyphrases
  • obstructive sleep apnea
  • ulcerative colitis
  • magnetic resonance
  • computed tomography
  • deep learning
  • magnetic resonance imaging
  • convolutional neural network
  • cystic fibrosis