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

This study evaluated the accuracy and reliability of artificial intelligence (AI)-reconstructed two-dimensional (2D) lateral cephalometric images. Researchers collected facial CT datasets from 40 patients, reformatted the original 1mm slices to 3mm, and then used an AI algorithm to restore them to 1mm resolution. Cephalometric images extracted from AI-reconstructed 1mm CT slices showed significantly reduced landmark discrepancies compared to those from 3mm CT slices (p<0.05), particularly at critical landmarks such as the sella, basion, and nasion. Interobserver and intraobserver reliability analyses demonstrated high consistency, with Cronbach’s α exceeding 0.7 in most cases. Aview Smart Slicer (Coreline Soft) was used to automatically reconstruct 2D lateral cephalometric images from CT scans, supporting AI-based image analysis.