切换至 "中华医学电子期刊资源库"

中华口腔医学研究杂志(电子版) ›› 2025, Vol. 19 ›› Issue (01) : 70 -74. doi: 10.3877/cma.j.issn.1674-1366.2025.01.010

综述

卷积神经网络辅助三维头影解剖标志自动化目标检测的研究与应用进展
庄文博1, 胡越1, 陈沁豪1, 商莉1, 张逸天2, 桂海军3,()   
  1. 1.上海交通大学医学院,上海 200025
    2.上海交通大学电子信息与电气工程学院,上海 200240
    3.上海交通大学医学院附属第九人民医院口腔颅颌面科,上海交通大学口腔医学院,国家口腔医学中心,国家口腔疾病临床医学研究中心,上海市口腔医学重点实验室,上海市口腔医学研究所,上海 200011
  • 收稿日期:2024-07-18 出版日期:2025-02-01
  • 通信作者: 桂海军
  • 基金资助:
    上海交通大学医学院大学生创新训练计划(1723X012)上海交通大学医学院附属第九人民医院研究型医师培育项目(2022hbyjxys-ghj)

Advances in research and application of convolutional neural network - assisted automated target detection of anatomical landmarks in three-dimensional cephalograms

Wenbo Zhuang1, Yue Hu1, Qinhao Chen1, Li Shang1, Yitian Zhang2, Haijun Gui3,()   

  1. 1.Shanghai Jiao Tong University School of Medicine,Shanghai 200025,China
    2.School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China
    3.Department of Oral and Cranio-Maxillofacial Surgery,Shanghai Ninth People' s Hospital,College of Stomatology,Shanghai Jiao Tong University;National Center for Stomatology,National Clinical Research Center for Oral Diseases,Shanghai Key Laboratory of Stomatology,Shanghai Research Institute of Stomatology,Shanghai 200011,China
  • Received:2024-07-18 Published:2025-02-01
  • Corresponding author: Haijun Gui
引用本文:

庄文博, 胡越, 陈沁豪, 商莉, 张逸天, 桂海军. 卷积神经网络辅助三维头影解剖标志自动化目标检测的研究与应用进展[J/OL]. 中华口腔医学研究杂志(电子版), 2025, 19(01): 70-74.

Wenbo Zhuang, Yue Hu, Qinhao Chen, Li Shang, Yitian Zhang, Haijun Gui. Advances in research and application of convolutional neural network - assisted automated target detection of anatomical landmarks in three-dimensional cephalograms[J/OL]. Chinese Journal of Stomatological Research(Electronic Edition), 2025, 19(01): 70-74.

基于二维影像的头影测量分析技术是当前临床诊断的“金标准”,但仍存在由于二维图像易失真、解剖标志重叠等导致的“解剖误差”和人工手动标点造成的“人工误差”。三维头影测量分析目前已广泛应用于临床诊断,对于“解剖误差”有着重要辅助作用;自动化头影测量分析是利用图像处理与深度学习对头影解剖标志进行自动识别与标点,对于改善二维测量分析的“人工误差”有重要意义。卷积神经网络(CNN)是目前基于深度学习最有效的图像处理与目标检测方法,在三维头影解剖标志自动化目标测量领域有着巨大应用潜力。本文综合国内外有关研究与应用,回顾了三维头影解剖标志测量分析体系的现状,并对基于CNN 的牙颌面三维头影解剖标志自动化目标检测的研究进展进行了探索与综述。

Cephalometric analysis technology based on two-dimensional images has always been the‘golden standard’.Still,there are the problems of‘anatomical errors’caused by the distortion of two-dimensional images and overlapping of anatomical landmarks,and‘artificial errors’caused by manual punctuation.Three-dimensional(3D)cephalometric analysis,which has been widely used in clinical diagnosis,and playing a more and more important role for resolving the‘anatomical error’problem.Automatic cephalometric analysis,which uses image processing and deep learning for identifying and punctuating cephalometric landmarks automatically,could be used for resolving the‘manual error’problem.Convolutional neural network(CNN)based on deep learning is currently the most effective technology of image processing and target detection,which has shown its great potential for automatic target detection of 3D cephalometric landmarks.Based on literature review,we summarized the current status of 3D cephalometric analysis and the research progress of CNN for automatic target detection of 3D cephalometric landmarks.

表1 三维头影解剖标志体系
研究者 研究内容 使用解剖点 筛选原因
Montúfar等[7] 头影特征点注释的混合算法探究 S(蝶鞍点)、N(鼻根点)、Ba(颅底点)、O(眶点)、ANS(前鼻棘点)、PNS(后鼻棘点)、A(上齿槽座点)、B(下齿槽座点)、Go(下颌角点)、Pg(颏前点)、Me(颏下点)、Po(耳点)、Gn(颏顶点)、L1(下切牙点)、U1(上切牙点) 判断由锥形束计算机断层扫描角度上出发的自动头影测量准确性,平均定位误差小
Jie等[8] 骨整形术后颧上颌区软硬组织变化的相关性研究 N、Ol(左眶点)、Or(右眶点)、Po、ANS、Me、Gol(左下颌角点)、Gor(右下颌角点)、Ba 研究结果具有明确统计学意义,相关点位检测价值高
Jeon等[9] 头影测量比较自动化分析 S、N、Po、O、Ar(关节点)、PNS、ANS、A、U1、U6(上颌第一磨牙点)、L1、L6(下颌第一磨牙点)、Go、B、Me 结果证明相关解剖点自动头影测量分析的临床诊断准确性高
Kang等[10] 三维头影特征点检测 bregma(前囟点)、N、center of foramen magnum(枕骨大孔中心点)、S、ANS、Pg、O、Po、mandibular foramen(下颌孔点)、mental foramen(颏孔点) 解剖标志点目标检测相对误差小,相关点位检测价值高
Sam等[11] 不同三维头影特征系统评价 S、Ba、N、ANS、A、B、Pg、Gn、Me 研究得出正中矢状面和双侧对称性的三维头影解剖标志有最高的可靠性
Jiang等[12] 纯磁共振成像(MRI)头影测量分析方案实现 S、N 涉及较少点位选择而需要宏观考量,丰富在有较多限制条件下的点位选用方案
表2 三维头影测量参数体系
研究者 研究内容 测量参数 筛选原因
Elshebiny等[13] 三维软组织预测精度研究 Franto-nasal angle(FNA)(鼻额角)、Naso-labial angle(NLA)(鼻唇角)、Labio-mental angle(LMA)(颏唇角)、A-A'(A-软组织A距离)、Is-U1(上唇厚度)、Li-L1(下唇厚度)、B-B'(B-软组织B距离)、Pog-Pog'(软组织颏厚度)、Gn-Gn'(Gn-软组织Gn距离)、N'-Tip of nose(鼻长度)、Sn-stms(上唇长度)、stmi-Me'(下唇长度)、P-Sn(鼻深度)、Subalar R.-Subalar L(. 鼻翼基底宽度)、Chelion L.-Chelion R(. 颏部宽度) 判断样本在手术治疗前后的软组织差异性,并得到了具有统计学意义的结果
Jodeh等[14] 二维和三维头影测量的比较 SNA(蝶鞍-鼻根-上颌前缘角)、SNB(蝶鞍-鼻根-下颌前缘角)、ANB(上颌前缘-鼻根-下颌前缘角)、SN-MP(颅底平面-下颌平面角)、MP-FH(下颌平面-眼耳平面角)、OP-FH(咬合平面-眼耳平面角)、OP-SN(咬合平面-颅底平面角)、PP-OP(腭平面-咬合平面角)、PP-MP(腭平面-下颌平面角)、U1-SN(上中切牙-前颅底平面角)、L1-MP(下中切牙-下颌平面角)、U1-L1(上下中切牙角) 显示出2D与3D头影测量参数统计学意义上的差异,具备相当选取意义
Oh等[15] 头影测量标志检测的深层解剖环境特征学习 ANB(上颌前缘-鼻根-下颌前缘角)、SNB(蝶鞍-鼻根-下颌前缘角)、SNA(蝶鞍-鼻根-上颌前缘角)、ODI(咬合深度指标)、ADPI(前后发育不良指标)、FHI(面部高度指数)、FMA(下颌平面角)、MW(改良wits参数) 围绕开牙合畸形等特定相关参数进行研究,有相当的特征性
Franke等[16] 下颌支的三维一致性研究 相关分割部分的体积与表面积,均方根误差、平均值与平均值的绝对距离和符号距离函数等有关参数 直接涉及高维度参数,将数据处理方式纳入考量,提供更为全面的选用方向
[1]
Leung MY,Leung YY.Three - dimensional evaluation of mandibular asymmetry: A new classification and three -dimensional cephalometric analysis[J].Int J Oral Maxillofac Surg,2018,47(8):1043-1051.DOI:10.1016/j.ijom.2018.03.021.
[2]
Oh J,Han JJ,Ryu SY,et al.Clinical and cephalometric analysis of facial soft tissue[J].J Craniofac Surg,2017,28(5):e431-e438.DOI:10.1097/scs.0000000000003614.
[3]
Santos RMG,de Martino JM,Haiter Neto F,et al.Cone beam computed tomography-based cephalometric norms for Brazilian adults[J].Int J Oral Maxillofac Surg,2018,47(1):64-71.DOI:10.1016/j.ijom.2017.06.030.
[4]
Liu K,Yingwang J,Zhang L,et al.A rare complication following anesthesia:Arytenoid dislocation during orthognathic surgery[J].J Oral Maxillofac Surg,2019,77(5):959-964.DOI:10.1016/j.joms.2018.11.029.
[5]
Ham YG,Kim JH,Luo JJ.Deep learning for multi-year ENSO forecasts[J].Nature,2019,573(7775):568-572.DOI:10.1038/s41586-019-1559-7.
[6]
Junaid N,Khan N,Ahmed N,et al.Development,application,and performance of artificial intelligence in cephalometric landmark identification and diagnosis:A systematic review[J].Healthcare(Basel),2022,10(12):2454.DOI:10.3390/healthcare10122454.
[7]
Montúfar J,Romero M,Scougall-Vilchis RJ.Hybrid approach for automatic cephalometric landmark annotation on cone-beam computed tomography volumes[J].Am J Orthod Dentofacial Orthop,2018,154(1):140-150.DOI:10.1016/j.ajodo.2017.08.028.
[8]
Jie B,Yao B,An J,et al.Correlation between soft and hard tissue changes in the zygomaticomaxillary region after bone contouring surgery for fibrous dysplasia:A preliminary study[J].J Oral Maxillofac Surg,2019,77(9):1904.e1-1904.e11.DOI:10.1016/j.joms.2019.05.002.
[9]
Jeon S,Lee KC.Comparison of cephalometric measurements between conventional and automatic cephalometric analysis using convolutional neural network[J].Prog Orthod,2021,22(1):14.DOI:10.1186/s40510-021-00358-4.
[10]
Kang SH,Jeon K,Kang SH,et al.3D cephalometric landmark detection by multiple stage deep reinforcement learning[J].Sci Rep,2021,11(1):17509.DOI:10.1038/s41598-021-97116-7.
[11]
Sam A,Currie K,Oh H,et al.Reliability of different threedimensional cephalometric landmarks in cone-beam computed tomography:A systematic review[J].Angle Orthod,2019,89(2):317-332.DOI:10.2319/042018-302.1.
[12]
Jiang X,Pei J,Liu J,et al.An MRI-only three-dimensional cephalometry protocol based on the integrated and modular architecture of the human head[J].Curr Med Imaging,2023.DOI:10.2174/0115734056258953231026094236.
[13]
Elshebiny T,Morcos S,Mohammad A,et al.Accuracy of threedimensional soft tissue prediction in orthognathic cases using dolphin three-dimensional software[J].J Craniofac Surg,2019,30(2):525-528.DOI:10.1097/scs.0000000000005037.
[14]
Jodeh DS,Kuykendall LV,Ford JM,et al.Adding depth to cephalometric analysis:Comparing two- and three-dimensional angular cephalometric measurements [J].J Craniofac Surg,2019,30(5):1568-1571.DOI:10.1097/scs.0000000000005555.
[15]
Oh K,Oh IS,Le VNT,et al.Deep anatomical context feature learning for cephalometric landmark detection [J].IEEE J Biomed Health Inform,2021,25(3):806-817.DOI:10.1109/jbhi.2020.3002582.
[16]
Franke A,Sequenc AF,Sembdner P,et al.Three-dimensional measurements of symmetry for the mandibular ramus[J].Ann Anat,2024,253:152229.DOI:10.1016/j.aanat.2024.152229.
[17]
Lecun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-22324.DOI:10.1109/5.726791.
[18]
Krizhevsky A,Sutskever I,Hinton GE.ImageNet classification with deep convolutional neural networks[J].Commun ACM,2017,60(6):84-90.DOI:10.1145/3065386.
[19]
Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition[J].Computer Science,2014.DOI:10.48550/arXiv.1409.1556.
[20]
Szegedy C,Liu W,Jia Y,et al.Going deeper with convolutions[C/OL]//2015 IEEE Conference on Computer Vision and Pattern Recognition(CVPR),Boston,2015:1-9.DOI:10.1109/CVPR.2015.7298594.
[21]
He K,Zhang X,Ren S,et al.Deep residual learning for image recognition[C/OL]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR),Las Vegas,2016:770-778.DOI:10.1109/CVPR.2016.90.
[22]
Kunz F,Stellzig - Eisenhauer A,Zeman F,et al.Artificial intelligence in orthodontics:Evaluation of a fully automated cephalometric analysis using a customized convolutional neural network[J].J Orofac Orthop,2020,81(1):52-68.DOI:10.1007/s00056-019-00203-8.
[23]
Kim MJ,Liu Y,Oh SH,et al.Evaluation of a multi-stage convolutional neural network-based fully automated landmark identification system using cone-beam computed tomographysynthesized posteroanterior cephalometric images[J].Korean J Orthod,2021,51(2):77-85.DOI:10.4041/kjod.2021.51.2.77.
[24]
Zhang J,Liu M,Wang L,et al.Context-guided fully convolutional networks for joint craniomaxillofacial bone segmentation and landmark digitization[J].Med Image Anal,2020,60:101621.DOI:10.1016/j.media.2019.101621.
[25]
Gil SM,Kim I,Cho JH,et al.Accuracy of auto-identification of the posteroanterior cephalometric landmarks using cascade convolution neural network algorithm and cephalometric images of different quality from nationwide multiple centers[J].Am J Orthod Dentofacial Orthop,2022,161(4):e361-e371.DOI:10.1016/j.ajodo.2021.11.011.
[26]
Lang Y,Lian C,Xiao D,et al.Localization of craniomaxillofacial landmarks on CBCT images using 3D Mask R-CNN and local dependency learning[J].IEEE Trans Med Imaging,2022,41(10):2856-2866.DOI:10.1109/tmi.2022.3174513.
[27]
张紫涵,熊鑫,王军.三维头影测量的研究现状和应用发展[J].国际口腔医学杂志,2020,47(6):739-744.DOI:10.7518/gjkq.2020092.
[28]
Lee JH,Yu HJ,Kim MJ,et al.Automated cephalometric landmark detection with confidence regions using Bayesian convolutional neural networks[J].BMC Oral Health,2020,20(1):270.DOI:10.1186/s12903-020-01256-7.
[29]
Zeng M,Yan Z,Liu S,et al.Cascaded convolutional networks for automatic cephalometric landmark detection[J].Med Image Anal,2021,68:101904.DOI:10.1016/j.media.2020.101904.
[30]
Weingart JV,Schlager S,Metzger MC,et al.Automated detection of cephalometric landmarks using deep neural patchworks[J].Dentomaxillofac Radiol,2023,52(6):20230059.DOI:10.1259/dmfr.20230059.
[1] 李洋, 蔡金玉, 党晓智, 常婉英, 巨艳, 高毅, 宋宏萍. 基于深度学习的乳腺超声应变弹性图像生成模型的应用研究[J/OL]. 中华医学超声杂志(电子版), 2024, 21(06): 563-570.
[2] 罗刚, 泮思林, 孙玲玉, 李志新, 陈涛涛, 乔思波, 庞善臣. 一种新型语义网络分析模型对室间隔完整型肺动脉闭锁和危重肺动脉瓣狭窄胎儿右心发育不良程度的评价作用[J/OL]. 中华医学超声杂志(电子版), 2024, 21(04): 377-383.
[3] 孔德铭, 刘铮, 李睿, 钱文伟, 王飞, 蔡道章, 柴伟. 人工智能辅助全髋关节置换三维术前规划准确性评价[J/OL]. 中华关节外科杂志(电子版), 2024, 18(04): 431-438.
[4] 张嘉炜, 王瑞, 张克诚, 易磊, 周增丁. 烧烫伤创面深度智能检测模型P-YOLO的建立及测试效果[J/OL]. 中华损伤与修复杂志(电子版), 2024, 19(05): 379-385.
[5] 林宥宏, 李运峰. 牵张成骨在牙颌面畸形治疗中的应用[J/OL]. 中华口腔医学研究杂志(电子版), 2025, 19(01): 9-15.
[6] 李明娟, 林陈心子, 吴姝萱, 万启龙. 合并牙列缺损的牙颌面畸形的病因、临床特点及诊疗策略[J/OL]. 中华口腔医学研究杂志(电子版), 2025, 19(01): 1-8.
[7] 叶莉, 杜宇. 深度学习在牙髓根尖周病临床诊疗中的应用[J/OL]. 中华口腔医学研究杂志(电子版), 2024, 18(06): 351-356.
[8] 张悦, 张可, 邓锶锶, 向青, 郭亚豪, 曹键, 罗涛, 孟占鳌. 深度学习图像重建三低方案在肾动脉血管成像中的应用[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2025, 19(01): 76-82.
[9] 黄俊龙, 李文双, 李晓阳, 刘柏隆, 陈逸龙, 丘惠平, 周祥福. 基于盆底彩超的人工智能模型在女性压力性尿失禁分度诊断中的应用[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 597-605.
[10] 犹成亿, 尤恒, 叶东樊, 张雯, 刘禹, 王仁宇, 苏琳茜, 甘慧, 徐智. 基于3D Res U-Net-Faster RCNN 技术和CT 影像学特征的肺结节性质预测模型的建立[J/OL]. 中华肺部疾病杂志(电子版), 2024, 17(05): 673-679.
[11] 赵毅, 李昶田, 唐文博, 白雪婷, 刘荣. 腹腔镜术中超声主胰管自动识别模型的临床应用[J/OL]. 中华腔镜外科杂志(电子版), 2024, 17(05): 290-294.
[12] 尹泽新, 杨继林, 李有尧, 吴美龙, 刘利平. 肝癌微血管侵犯的术前预测研究进展[J/OL]. 中华肝脏外科手术学电子杂志, 2025, 14(01): 128-134.
[13] 胡师尧, 陈媛媛, 李辰, 严宏. 深度学习在后发性白内障混浊分析中的应用研究[J/OL]. 中华眼科医学杂志(电子版), 2024, 14(05): 262-268.
[14] 潘清, 葛慧青. 基于机械通气波形大数据的人机不同步自动监测方法[J/OL]. 中华重症医学电子杂志, 2024, 10(04): 399-403.
[15] 孙铭远, 褚恒, 徐海滨, 张哲. 人工智能应用于多发性肺结节诊断的研究进展[J/OL]. 中华临床医师杂志(电子版), 2024, 18(08): 785-790.
阅读次数
全文


摘要