Novel prediction models for pharyngeal-airway volume based on the cranial-base and midsagittal cross-sectional area of the airway in the pharyngeal region: A cephalometric and magnetic resonance imaging study

Objective: The objective of the study was to elucidate the association between cranial base (Bjork-Jarabak analysis), midsagittal cross-sectional area of the airway in the pharyngeal region (MCSA-PR) data and pharyngeal-airway volume (PAV) and develop a model that could help clinicians predict PAV using two-dimensional (2D) data (Bjork polygon and MCSA-PR). Materials and Methods: Pre-treatment lateral cephalometric radiographs and magnetic resonance imaging (MRI) scans of 82 women were categorized into three anteroposterior skeletal groups based on ANB angle: Class I (n = 29), 1.5° ≤ ANB≤5.1°; Class II (n = 26), ANB >5.1°; Class III (n = 27), ANB <1.5°. The Bjork polygon, MCSA-PR data from cephalograms and PAV data from MRI scans were examined. Intergroup comparisons were performed using the Kruskal–Wallis test and one-way analysis of variance (ANOVA), with pairwise comparisons conducted using the Bonferroni-corrected Mann–Whitney U-test for the Kruskal–Wallis test and Bonferroni-corrected multiple comparison test for one-way ANOVA. Forward multiple linear regression was used to create model equations for predicting PAV. Results: MCSA-PR and anterior (N-S) and posterior (S-Ar) cranial-base lengths were positively correlated with the PAV. We developed four models; three operated at the group level, and one encompassed the entire sample. Notably, all models could effectively explain the variance in the PAV data. The model for the Class I group was the strongest (adjusted R² = 0.77). Conclusion: Our findings indicate the remarkable potential of the MCSA-PR, N-S and Bjork sum angles (BSA) as predictors of the PAV and the relevance of 2D cephalometric and cranial-base parameters in predicting the three-dimensional (3D) pharyngeal-airway size.