上海口腔医学

上海口腔医学 ›› 2022, Vol. 31 ›› Issue (5): 454-459.doi: 10.19439/j.sjos.2022.05.002

• 论著 • 上一篇    下一篇

U-Net神经网络分割锥形束CT影像中下颌磨牙牙体与牙髓腔及其准确性验证

林翔1, 傅裕杰1, 任根强2, 温佳欢1, 陈宇飞2,*, 张旗1,*   

  1. 1.同济大学口腔医学院·同济大学附属口腔医院 牙体牙髓病科,上海牙组织修复与再生工程技术中心,上海 200072;
    2.同济大学 电子与信息工程学院,上海 201804
  • 收稿日期:2021-03-18 修回日期:2021-05-10 出版日期:2022-10-25 发布日期:2022-11-01
  • 通讯作者: 陈宇飞,E-mail: yufeichen@tongji.edu.cn;张旗,E-mail: qizhang@tongji.edu.cn。*共同通信作者
  • 作者简介:林翔(1994-),男,硕士,E-mail：1731270@tongji.edu.cn
  • 基金资助:
    国家自然科学基金（81870760）; 中央基本高校经费（22120190217）; 上海市卫生健康委员会卫生行业临床研究专项（202040282）; 上海申康医院发展中心临床三年行动计划资助（SHDC2020CR3058B）

Segmentation, accuracy validation of mandibular molar, pulp cavity on cone-beam CT images by U-net neural network

LIN Xiang1, FU Yu-jie1, REN Gen-qiang2, WEN Jia-huan1, CHEN Yu-fei2, ZHANG Qi1   

  1. 1. Department of Endodontics, School and Hospital of Stomatology, Tongji University, Shanghai Engineering Research Center of Tooth Restoration and Regeneration. Shanghai 200072;
    2. College of Electronics and Information Engineering, Tongji University. Shanghai 201804, China
  • Received:2021-03-18 Revised:2021-05-10 Online:2022-10-25 Published:2022-11-01

PDF (PC)

298

摘要： 目的: 使用U-net卷积神经网络实现锥形束CT（cone-beam CT, CBCT）影像中下颌磨牙的牙体和牙髓腔的自动分割,采用基于显微CT（Micro-CT）扫描结果构建的三维模型作为金标准,评估分割准确性。方法: 从同济大学附属口腔医院放射科收集20组包含完整单侧下颌磨牙的口腔小视野CBCT数据,预处理后,由牙体牙髓病学专家使用MITK Workbench软件手动标注牙体与牙髓腔,作为U-net神经网络分割算法的训练集。另收集5颗下颌磨牙和相应的小视野CBCT数据,5组数据经相同预处理后作为测试集。随后由完成训练的神经网络和同一专家对测试集数据进行牙体和牙髓腔分割和三维重建。离体牙预处理后行Micro-CT扫描,将三维重建后获得的模型作为金标准。分别比较测试集数据中,专家的手动标注、神经网络分割结果与金标准两两之间的差异。采用Dice相似性系数（Dice similarity coefficient, DSC）、平均对称表面距离（average symmetric surface distance, ASSD）、Hausdorff距离（Hausdorff distance, HD）和形态差异分析对结果进行评估。采用SPSS 20.0软件包对数据进行统计学分析。结果: 神经网络分割结果与金标准相比,其牙体组的DSC为(95.30±1.01)%、ASSD为(0.11±0.02) mm、HD为(1.05±0.31) mm,牙髓腔组的DSC为(81.21±2.27)%、ASSD为(0.15±0.05) mm、HD为(3.29±1.85) mm,结合形态差异分析结果显示,神经网络的分割结果与金标准的牙体与髓室部分基本相似,但在根管部分,能分割出较粗的根管,对于根管下段和侧支根管等较细的根管分割能力有限。结论: 在现有实验条件下,以专家手动标注作为训练样本的U-net神经网络,实现了在CBCT影像上对下颌磨牙牙体与髓室的自动化精准分割。但对根管部分,其分割结果有待进一步提升。

关键词: U-net卷积神经网络, 锥形束CT, 显微CT, 图像分割, 下颌磨牙

Abstract: PURPOSE: To realize the automatic segmentation of mandibular molar and pulp cavity on cone-beam CT (CBCT) images by U-net convolutional neural network, and to use the 3D models reconstructed by Micro-CT data as the ground truth to validate its accuracy. METHODS: Twenty groups of small field of view(FOV) CBCT data containing complete mandibular molars were collected from the Department of Radiology, Affiliated Stomatology Hospital of Tongji University. After preprocessing, an endodontic specialist labeled teeth and pulp cavities by MITK Workbench software. These data were used as the training set for training U-net neural network. In addition, five mandibular molars and corresponding small FOV CBCT data were collected. These five CBCT were processed in the same way and used as the testing set. Then, teeth and pulp cavities on CBCT images of the testing set were segmented and reconstructed by U-net neural network and the same specialist. The isolated teeth were scanned by a Micro-CT machine after preprocessing and the results were reconstructed to 3D models, which were used as the ground truth. Then the 3D models reconstructed by the specialist’s labeling, U-net network segmentation results, and the ground truth in the testing set were compared. Dice similarity coefficient(DSC), average symmetric surface distance (ASSD), Hausdorff distance (HD), and morphological analysis were used to evaluate the results. SPSS 20.0 software package was used for statistical analysis. RESULTS: Compared with the ground truth, the segmentation accuracy of the U-net neural network measured by DSC, ASSD, and AHD was (95.30±1.01)%, (0.11±0.02) mm, and (1.05±0.31) mm in teeth and (81.21±2.27)%, (0.15±0.05) mm, and (3.29±1.85) mm in the pulp cavity, respectively. Morphological analysis results showed that the U-net network segmentation results were similar to the ground truth in tooth and pulp chamber. As for the segmentation results of root canals, only thick root canals could be segmented rather than the thin root canals, such as the canals in the apical third and lateral root canals. CONCLUSIONS: Under the experimental conditions, the U-net neural network trained by the specialist’s labeling realized the automatic and accurate segmentation of mandibular molar and their pulp chamber on CBCT images. For the segmentation of root canals, the results need to be further improved.

Key words: U-net convolutional neural network, Cone-beam CT, Micro-CT, Image segmentation, Mandibular molars

中图分类号: