UCLA

Cleft lip with or without cleft palate (CL/P) is one of the most common craniofacial anomalies. Patients with CL/P often require extensive and prolonged orthodontic treatment due the skeletal and dental malocclusions that frequently manifest, making the orthodontic burden of care significant compared to the non-cleft patient. Orthodontic treatment may be required at various stages in their dental and skeletal development, during which orthodontic records must regularly be taken, including cephalometric radiographs. These radiographs allow for analysis of skeletal and dental relationships by relating various anatomical landmarks through linear and angular measurements.

Due to the time-consuming nature of the current computer-aided method, which currently involves an orthodontist locating points of a lateral skull radiograph on a computer monitor, systems have recently been developed to automate the cephalometric process. However, while systems such as CephX have been shown to significantly shorten analyzing time, the accuracy of the measurements is inadequate. Further, radiographs of CL/P-affected patients are often excluded from the sample, as identification of cephalometric landmarks is more complicated due to abnormal anatomy. There is promise that lateral simulated 2D cephalometric projections from CBCTs improve the accuracy of cephalometric measurements over 2D cephalograms.

This study sought to compare the accuracy and analyzing time between web-based fully automated and computer-aided cephalometric analysis of lateral cephalometric images derived from cone-beam computed tomography in unilateral cleft lip and palate-affected patients. Both methods of cephalometric analysis were performed on 36 CBCTs obtained of individuals with unilateral cleft lip and palate in the mixed dentition stage.

Of the 12 measurements obtained, 4 measurements, U1-PP (°), SN-MP (°), U1-SN (°) and U1-NA (mm) were both statistically significant (P<0.05) and had mean differences above the clinically acceptable limit of 2 mm or 2°. The agreement interval fell outside the range of clinical acceptability for every measurement. Results also showed that the automated program took significantly longer than the computer-assisted method to produce a cephalometric analysis.

Based on these results, it is advisable to use CephX for cephalometric analysis in patients with cleft lip and palate only with clinician supervision and intervention. Further development is needed, particularly with regard obtaining measurements involving landmarks that are challenging to identify and those with multiple definitions before completely replacing computerized tracing.