Cone beam computed tomography registration for 3-D airway analysis based on anatomic landmarks

Objectives. Explore craniofacial landmarks reliability to superimpose cone beam computed tomography (CBCT) images and assess impact of plane reorientation on airway parameters. Study design. 10 CBCTs were marked 3 times at baseline, 3 T1, to test landmark reliability. Measurement errors of new coordinate system were tested using 3 T1, and other 10 paired CBCT images, at T1 and T2. Effect on upper airway was assessed using volume, surface area, and point-based analysis. Results. Tips of nasal bone and clivus and foramina spinosa defined the new coordinate system. Plane reorientation didn't affect landmark identification reliability and significantly reduced interlandmark distances from T1-T2. Airway volume changed by 25.76 +/- 24.9%, surface area by 13.85 +/- 10.8%, and mean part analysis was 0.43 +/- 0.3 mm. Strong correlation (R-2 > 65%) was found between airway analysis and large distances in second and third cervical vertebrae. Conclusions. Coordinate transformation significantly reduced measurement errors in longitudinal CBCT data; however, it is not designed to correct for large neck flexion.