Three-dimensional computed tomographic analysis of airway anatomy

Abstract

Purpose: To test the reliability of a 3-dimensional computed tomographic (3D-CT) analysis of airway size and shape and to correlate the 3D-CT findings with lateral cephalometric measurements.

Materials and methods: Fifteen pairs of preoperative maxillofacial 3D-CT scans and digital lateral cephalograms of patients treated for obstructive sleep apnea were used in the present study. Digital 3D-CT reconstructions were created and 12 measurements of airway size and 4 of shape were analyzed. The posterior airway space (PAS), middle airway space, and hyoid to mandibular plane distances were measured on the cephalograms. We then randomly selected 5 CT scans and 5 cephalograms which were analyzed blindly on 5 separate occasions by 2 investigators (Z.A., J.T.) to establish the intraclass correlation coefficients for inter- and intraexaminer reliability. All 15 pairs of images were used to compute the Pearson correlation coefficients to establish the relationship between the CT and cephalometric measurements.

Results: The intra- and interexaminer reliabilities were high for all CT (0.86 to 1.0 and 0.89 to 1.0, respectively; P < .001) and cephalometric measurements (0.84 to 1.0 and 0.91 to 0.99, respectively; P < .001). The CT measurements retroglossal anteroposterior dimension and distance between the genial tubercle and hyoid exhibited a positive correlation with the PAS (r = .60, P = 02 and r = .54, P = .04, respectively), and the lateral/anteroposterior dimension demonstrated an inverse correlation (r = -.68, P = .01) with the PAS.

Conclusion: The results of the present study indicate that the 3D-CT and lateral cephalometric measurements we selected are reliable and reproducible. The only cephalometric measurement that exhibited any correlation with the CT parameters was PAS.