AI-Assisted Segmentation and 3D Morphometric Characterization of the Edentulous Mandible from CBCT Data

This study presents the first structured CBCT-based dataset dedicated exclusively to the three-dimensional morphometric characterisation of the edentulous mandible. Cone-beam computed tomography (CBCT) scans were retrospectively collected from two imaging centres, and fully edentulous mandibles were identified and segmented using Dental Segmentator in 3D Slicer. The mandible and mandibular nerve were exported as STL files, processed in Meshmixer to generate standardised cross-sectional slices, and measured in MeshLab for ridge height, canal position, buccolingual width, and mandibular angulation. The resulting dataset captures significant anatomical variability among edentulous mandibles, reflecting the classical patterns of post-extraction bone resorption described by Atwood and Wical. The dataset includes original CBCT volumes, 3D models, morphometric measurement outputs, and metadata. This resource supports the development of implant planning tools, surgical simulations, mandibular biomechanical research, and AI-based anatomical modelling.