Adults' dental cone beam computed tomography images dataset for detecting and classifying missing teeth.
Authors
Affiliations (5)
Affiliations (5)
- Department of Dentistry, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310018, China.
- School of Cyberspace, Hangzhou Dianzi University, No. 1158, Ave. 2, Qiantang District, Hangzhou, Zhejiang, 310018, China.
- China Mobile (Zhejiang) Research & Innovation Institute, No. 19 Jiefang East Road, Shangcheng District, Hangzhou, Zhejiang, 310016, China.
- Innovation Center for Electronic Design Automation Technology, Hangzhou Dianzi University, Hangzhou, 310018, China. [email protected].
- Department of Dentistry, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310018, China. [email protected].
Abstract
The scarcity of high-quality Cone Beam Computed Tomography (CBCT) datasets with Three-Dimensional (3D) annotations hinders advancements in automated dental implant planning. Existing datasets lack precise 3D localization of tooth loss sites, particularly in complex cases presenting metal artifacts or existing implants. To address this, we curated a CBCT dataset from 158 patients across three institutions. Cases were rigorously screened for image quality (scores 2-4) and annotated using 3D Slicer, with artifact characteristics explicitly delineated. The dataset comprises a total of 158 CBCT volumes featuring 501 missing tooth sites, of which 85 volumes containing 114 specific sites were selected for 3D annotation (4,994 slices annotated). This resource provides 3D annotations of missing teeth to facilitate the training of deep learning models for automated implant planning. Furthermore, we established comprehensive baselines through clinical evaluations and deep learning modeling. These publicly available data aim to foster artificial intelligence (AI)-driven dental rehabilitation research.