Prediction of airway volume from lateral cephalograms and correlation among 2D and 3D measurements: A preliminary study

Abstract

Purpose

To correlate 3D airway (Vol_Ro) with 2D (Area_ImJ) and distances by public-program (ImageJ: ImJ) and commercial Romexis:Ro, (Area_Ro), to find the correlation coefficient (R) and the coefficient of determination (R²) values, to develop prediction equations from 2D to yield 3D.

Materials and methods

The sample comprised 17 subjects with 2D from cephalograms and reconstructed 2D and 3D from CBCT. They were divided into three subgroups according to Area ImJ: subgroup one: small, six patients (<650 mm², 15–35 years); subgroup two: medium, five patients (650–800 mm², 14–33 years); subgroup three: large, six patients (>800 mm², 19–34 years). The R and R² values were calculated. Linear regression analysis was evaluated to propose prediction models of 2D and 3D values.

Results

ICC showed an extremely high value of 0.992. R values of Area_ImJ and Vol_Ro in subgroup two, together with Area_Ro and Vol_Ro in subgroup three significantly presented extra high and very high correlations (R = 0.953, 0.919 and 0.855, respectively). R² between Area_ImJ and Vol_Ro showed the extra high and moderately high values as 0.908 and 0.731 in subgroups two and three, respectively. Also R² among distances in the frontal and lateral views for subgroup one and Vol_Ro exhibited 0.984 and 0.798. Prediction equations were developed from 2D distances and area to yield volumes. Calculated subgroup sample size was acceptable.

Conclusion

R and R² from the distances and area with volume for subgroups demonstrated moderately high to extremely high values. Predicton equations of volume were proposed in these subgroups from 2D.

Keywords:

• Prediction
• 2D–3D
• Correlation
• Airway

Acknowledgement

The authors thank the Planmeca Company, Finland for their support of the demonstration program (Romexis).