Evaluation of the clinical utility of lateral cephalometry reconstructed from computed tomography extracted by artificial intelligence

This study evaluated the clinical utility and accuracy of lateral cephalometric images generated from facial CT scans reconstructed from 3 mm slices to 1 mm using artificial intelligence (AI). Using data from 40 patients, the AView Smart Slicer (developed by Coreline Soft) was employed to reconstruct the images. Sixteen craniofacial landmarks were traced manually and using autotracing by three observers. Discrepancies between landmarks were assessed by comparing AI-reconstructed images to the original 1 mm CT and 3 mm CT images. The results showed that cephalometric images reconstructed with AI were significantly closer to the original 1 mm images than the conventional 3 mm images, particularly for critical landmarks such as the sella and basion. High interobserver and intraobserver reliability was demonstrated.