使用具有双嵌入模块的图卷积神经网络进行正颌外科手术规划：多医院数据集的外部验证
Orthognathic surgical planning using graph CNN with dual embedding module: External validations with multi-hospital datasets.
作者信息
Kim In-Hwan, Kim Jun-Sik, Jeong Jiheon, Park Jae-Woo, Park Kanggil, Cho Jin-Hyoung, Hong Mihee, Kang Kyung-Hwa, Kim Minji, Kim Su-Jung, Kim Yoon-Ji, Sung Sang-Jin, Kim Young Ho, Lim Sung-Hoon, Baek Seung-Hak, Kim Namkug
机构信息
Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea.
Department of Convergence Medicine, University of Ulsan, College of Medicine, Asan Medical Center, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 05505, South Korea.
出版信息
Comput Methods Programs Biomed. 2023 Dec;242:107853. doi: 10.1016/j.cmpb.2023.107853. Epub 2023 Oct 8.
BACKGROUND AND OBJECTIVE
Despite recent development of AI, prediction of the surgical movement in the maxilla and mandible by OGS might be more difficult than that of tooth movement by orthodontic treatment. To evaluate the prediction accuracy of the surgical movement using pairs of pre-(T0) and post-surgical (T1) lateral cephalograms (lat-ceph) of orthognathic surgery (OGS) patients and dual embedding module-graph convolution neural network (DEM-GCNN) model.
METHODS
599 pairs from 3 institutions were used as training, internal validation, and internal test sets and 201 pairs from other 6 institutions were used as external test set. DEM-GCNN model (IEM, learning the lat-ceph images; LTEM, learning the landmarks) was developed to predict the amount and direction of surgical movement of ANS and PNS in the maxilla and B-point and Md1crown in the mandible. The distance between T1 landmark coordinates actually moved by OGS (ground truth) and predicted by DEM-GCNN model and pre-existed CNN-based Model-C (learning the lat-ceph images) was compared.
RESULTS
In both internal and external tests, DEM-GCNN did not exhibit significant difference from ground truth in all landmarks (ANS, PNS, B-point, Md1crown, all P > 0.05). When the accumulated successful detection rate for each landmark was compared, DEM-GCNN showed higher values than Model-C in both the internal and external tests. In violin plots exhibiting the error distribution of the prediction results, both internal and external tests showed that DEM-GCNN had significant performance improvement in PNS, ANS, B-point, Md1crown than Model-C. DEM-GCNN showed significantly lower prediction error values than Model-C (one-jaw surgery, B-point, Md1crown, all P < 0.005; two-jaw surgery, PNS, ANS, all P < 0.05; B point, Md1crown, all P < 0.005).
CONCLUSION
We developed a robust OGS planning model with maximized generalizability despite diverse qualities of lat-cephs from 9 institutions.
背景与目的
尽管人工智能近年来有所发展，但通过口腔颌面外科手术（OGS）预测上颌骨和下颌骨的手术移动可能比正畸治疗中的牙齿移动更困难。本研究旨在使用正颌外科（OGS）患者术前（T0）和术后（T1）的头颅侧位片（lat-ceph）以及双嵌入模块-图卷积神经网络（DEM-GCNN）模型，评估手术移动的预测准确性。
方法
来自3个机构的599对数据用作训练集、内部验证集和内部测试集，来自其他6个机构的201对数据用作外部测试集。开发了DEM-GCNN模型（IEM，用于学习头颅侧位片图像；LTEM，用于学习标志点），以预测上颌骨中前鼻棘（ANS）和后鼻棘（PNS）以及下颌骨中B点和下颌第一磨牙牙冠（Md1crown）的手术移动量和方向。比较了OGS实际移动的T1标志点坐标（真实值）与DEM-GCNN模型预测值以及基于卷积神经网络的现有模型C（学习头颅侧位片图像）预测值之间的距离。
结果
在内部和外部测试中，DEM-GCNN在所有标志点（ANS、PNS、B点、Md1crown，所有P>0.05）与真实值之间均未表现出显著差异。当比较每个标志点的累积成功检测率时，DEM-GCNN在内部和外部测试中的值均高于模型C。在展示预测结果误差分布的小提琴图中，内部和外部测试均显示，DEM-GCNN在PNS、ANS、B点、Md1crown方面的性能比模型C有显著提高。DEM-GCNN的预测误差值显著低于模型C（单颌手术，B点、Md1crown，所有P<0.005；双颌手术，PNS、ANS，所有P<0.05；B点、Md1crown，所有P<0.005）。
结论
我们开发了一个强大的OGS规划模型，尽管来自9个机构的头颅侧位片质量各异，但该模型具有最大的通用性。
Zotero 插件