跳到正文
原文
Nature 健康护理主题·· 4 天前精选AI 评分64

DentAll框架:全景牙片中的上下文感知牙齿分割与缺牙定位研究

Context-aware tooth instance segmentation and missing-tooth localization in panoramic radiographs

AI 导读

一项发表于Scientific Reports的研究提出DentAll框架,集成牙齿实例分割与解剖学缺失位点预测。在1,325例全景牙片中,实例级分割达DSC 88.21%(匹配实例92.74%)、AP50为97.34%;缺牙定位器MTLP将IoU≥0.5达标率由仅用索引的25.20%提升至72.58%,无需显式缺牙标注即可利用同弓邻近牙几何与目标牙身份进行坐标回归预测。

推荐理由

研究团队提出DentAll框架,在全景牙片中实现牙齿实例分割与缺牙定位,为口腔影像AI应用提供可参考的技术路径。

正文 · 原文

We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.

Abstract

Reliable localization and segmentation of individual teeth are fundamental for automated dental diagnostics and computer-aided treatment planning. However, panoramic radiographs pose persistent challenges due to anatomical variability, overlapping structures, and inconsistent image quality, thereby keeping clinical workflows heavily dependent on manual interpretation. This study introduces DentAll, a framework that integrates tooth instance segmentation with contextual estimation of anatomically expected missing-tooth locations. The proposed segmentation with Contextual Spatial Attention Blocks (S-CSAB) integrates dilated convolutions and ROI-based positional encoding to enhance fine-grained boundary delineation while maintaining anatomical coherence across adjacent teeth. Building upon these contextual features, the Missing-Tooth Localization Predictor (MTLP) estimates missing-tooth locations via patient-conditioned coordinate regression, using the target tooth’s identity together with the geometry of its nearest same-arch neighboring teeth, without requiring explicit missing-tooth annotations. DentAll was evaluated on 1,325 panoramic radiographs using instance-level segmentation metrics, COCO-style mask average precision, comparative baselines, and controlled pseudo-missing localization experiments. Under an all-ground-truth evaluation protocol, the framework achieved a DSC of 88.21% (92.74% among successfully matched instances), AP50 of 97.34%, AP75 of 87.20%, and mAP50:95 of 68.58%, while contextual MTLP localization improved the IoU ≥ 0.5 rate from 25.20% (index-only formulation) to 72.58%. These findings demonstrate that integrating instance segmentation with patient-specific anatomical context provides an effective and computationally practical framework for structured analysis of panoramic dental radiographs under anatomical variability.

Funding

This work was supported by the National Research Foundation of Korea (NRF) under the Glocal R&D Support Program (RS-2026-25570738), funded by the Ministry of Education (MOE), Republic of Korea, and by an NRF grant funded by the Korea government (MSIT) (No. RS-2023-00217471).

Author information

Authors and Affiliations

  1. Department of Computer Science, Katholieke Universiteit (KU), Leuven, 3000, Belgium

    Shaily Bajpai

  2. Department of Industrial Design, Eindhoven University of Technology, Eindhoven, 5612 AE, Netherlands

    Hamza Shafiq

  3. Energy AI, Korea Institute of Energy Technology (KENTECH), Naju, 58330, South Korea

    Bumshik Lee

Authors

  1. Shaily Bajpai
  2. Hamza Shafiq
  3. Bumshik Lee

Corresponding author

Correspondence to Bumshik Lee.

Ethics declarations

Competing interest

The authors declare no competing interest.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bajpai, S., Shafiq, H. & Lee, B. Context-aware tooth instance segmentation and missing-tooth localization in panoramic radiographs. Sci Rep (2026). https://doi.org/10.1038/s41598-026-78690-0

Download citation

  • Received: 21 April 2026

  • Accepted: 27 October 2026

  • Published: 16 November 2026

  • DOI: https://doi.org/10.1038/s41598-026-78690-0

Keywords

来源:Nature 健康护理主题 · nature.com