文献速递牙周病口腔修复与种植口腔正畸

深度学习在牙周超声图像中分割釉质、牙骨质、牙槽骨与牙龈

Deep learning-based segmentation of enamel, cementum, alveolar bone, and gingiva in periodontal ultrasound images.
PubMed 文献 · J Dent · 2026-04-18

本研究开发了一种双流深度学习架构,用于牙周超声图像中釉质、牙骨质、牙槽骨和牙龈的四类同时分割,并定位釉牙骨质界、牙槽嵴顶和牙龈边缘。模型在752张内部图像训练、111张外部图像测试,分割中位Dice系数内部0.891、外部0.841,标志点定位中位距离误差0.06–0.08 mm。

要点

  • 模型可同时分割釉质、牙骨质、牙槽骨和牙龈四类牙周结构。
  • 外部测试中位Dice系数为0.841,泛化性能较好。
  • 釉牙骨质界、牙槽嵴顶和牙龈边缘定位中位误差0.06–0.08 mm。

核心结论

  • 所提出的深度学习模型能准确自动分割牙周超声图像中的多类结构,并基于分割掩膜高精度定位解剖标志点。

研究设计与样本

  • 研究提出一种具有随机块混洗特征的双流深度学习架构。
  • 模型在752张内部图像数据集上训练进行四类同时分割,并在111张外部测试集上验证。

主要发现

  • 内部集分割中位Dice相似系数、交并比、精确率、敏感度、95%豪斯多夫距离和平均对称表面距离分别为0.891、0.805、0.887、0.909、0.083 mm和0.028 mm。
  • 外部集分割中位Dice相似系数、交并比、精确率、敏感度、95%豪斯多夫距离和平均对称表面距离分别为0.841、0.728、0.781、0.921、0.089 mm和0.032 mm。
  • 釉牙骨质界、牙槽嵴顶和牙龈边缘的定位中位距离误差分别为0.06 mm、0.08 mm和0.06 mm。

临床怎么用

  • 该自动多类分割模型可辅助口腔临床医生可视化和解读牙周超声图像。
  • 该方法有望支持超声在牙周状况评估及术前数字化规划中的更广泛临床应用,包括牙周病管理、修复治疗和正畸护理。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVES: To develop a deep learning-based multi-class segmentation model for the simultaneous segmentation of key periodontal structures, including enamel, cementum, alveolar bone, and gingiva, in ultrasound images, and to enable precise localization of the cementoenamel junction (CEJ), alveolar bone crest (ABC), and gingival margin (GM). METHODS: A novel dual-stream deep learning architecture featuring stochastic block shuffling was proposed. The model was trained for simultaneous four-class segmentation on an internal dataset of 752 images and validated on an external test set of 111 images. The resulting segmentation masks were subsequently used to identify three anatomical landmarks: the CEJ, ABC, and GM. RESULTS: The model demonstrated strong segmentation performance, with median Dice similarity coefficient, intersection over union, precision, sensitivity, 95% Hausdorff distance, and average symmetric surface distance values of 0.891, 0.805, 0.887, 0.909, 0.083 mm, and 0.028 mm, respectively, for the internal set, and 0.841, 0.728, 0.781, 0.921, 0.089 mm, and 0.032 mm, respectively, for the external set. In the assessment of landmark localization accuracy, the model achieved median distance errors of 0.06 mm, 0.08 mm, and 0.06 mm for the CEJ, ABC, and GM, respectively. CONCLUSION: The proposed deep learning model enabled accurate automated multi-class segmentation of periodontal structures in ultrasound images and facilitated highly precise localization of anatomical landmarks derived from the segmentation masks. CLINICAL SIGNIFICANCE: The proposed automatic multi-class segmentation model may assist dental clinicians in visualizing and interpreting periodontal ultrasound images. This approach shows promise for supporting broader clinical adoption of ultrasonography for the evaluation of periodontal conditions and preoperative digital planning, including periodontal disease management, restorative treatment, and orthodontic care.

来源:J Dent,查看原文。版权归原作者所有。

引用

  1. Piao JZ, Hu KS, Jung HS, et al. Deep learning-based segmentation of enamel, cementum, alveolar bone, and gingiva in periodontal ultrasound images. J Dent. 2026. PMID: 42009189. doi:10.1016/j.jdent.2026.106705

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。