上海交通大学学报 ›› 2022, Vol. 56 ›› Issue (11): 1509-1517.doi: 10.16183/j.cnki.jsjtu.2021.103

• 生物医学工程 • 上一篇    下一篇

基于点云语义分割算法的下颌角截骨面设计

吕超凡1, 言颖杰2, 林力2,3, 柴岗2, 鲍劲松1()   

  1. 1.东华大学 机械工程学院,上海 201620
    2.上海交通大学医学院附属第九人民医院 整复外科, 上海 200011
    3.上海交通大学 塑性成形技术与装备研究院,上海 200030
  • 收稿日期:2021-04-05 出版日期:2022-11-28 发布日期:2022-12-02
  • 通讯作者: 鲍劲松 E-mail:bao@dhu.edu.cn
  • 作者简介:吕超凡(1997-),男,安徽省黄山市人,硕士生,从事三维深度学习研究.
  • 基金资助:
    上海申康医院发展中心临床三年行动计划(SHDC2020CR3070B);上海市科学技术委员会项目(18DZ2201900);上海市科学技术委员会项目(19441912300)

Design of Mandibular Angle Osteotomy Plane Based on Point Cloud Semantic Segmentation Algorithm

LÜ Chaofan1, YAN Yingjie2, LIN Li2,3, CHAI Gang2, BAO Jinsong1()   

  1. 1. College of Mechanical Engineering, Donghua University, Shanghai 201620, China
    2. Department of Plastic and Reconstructive Surgery, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200011, China
    3. Institute of Forming Technology and Equipment, Shanghai Jiao Tong University, Shanghai 200030, China
  • Received:2021-04-05 Online:2022-11-28 Published:2022-12-02
  • Contact: BAO Jinsong E-mail:bao@dhu.edu.cn

摘要:

下颌角截骨手术是近年来较为热门的颅面整形手术.现阶段,下颌角截骨的术前方案设计通常由具有一定年资的医生完成,过程繁琐且耗时较长.为了提高截骨手术术前规划效率,提出一种基于点云语义分割网络的下颌角截骨面设计方法.对颅骨电子计算机断层扫描(CT)数据进行三维重建和表面点采样,将下颌骨三维模型转换为点云数据,然后通过基于Transformer的点云语义分割网络预测下颌骨点云中的截骨区域,最后根据点云分割结果计算出下颌角截骨平面.所提网络主要包括两个部分:一是基于注意力机制的本地特征提取层,用于提取细粒度局部结构信息;二是基于Transformer的非本地特征提取层,用于提取点云的全局上下文信息.在构建的下颌骨语义分割数据集上,将所提算法与其他点云语义分割算法进行比较.结果表明:所提算法能实现最佳的下颌角截骨区域预测,优于目前常见的点云语义分割算法.

关键词: 下颌角截骨术, 点云分割, 自注意力, 深度学习

Abstract:

Mandibular angle osteotomy is a popular craniofacial plastic surgery in recent years. Usually, preoperative planning of mandibular angle osteotomy is completed by an experienced doctor, which is cumbersome and time-consuming. In order to improve the efficiency of osteotomy planning, a design method of mandibular angle osteotomy plane based on point cloud semantic segmentation network is proposed. After three-dimensional reconstruction of the skull computer tomography (CT) scan data, the three-dimensional model of the mandible is converted into point cloud data through uniform sampling. The resection area of the mandible is predicted by the proposed algorithm, which is used to calculate the mandibular angle osteotomy plane. The proposed semantic segmentation network mainly includes 2 parts: a local feature extraction layer based on the attention mechanism, which is used to extract fine-grained local structure information, and a non-local feature extraction layer based on Transformer, which is used to extract the global context information of the point cloud. On the constructed mandible semantic segmentation data set, the proposed algorithm is compared with other point cloud semantic segmentation algorithms. The results show that the proposed algorithm can achieve the best prediction of the mandibular angle resection area, which is better than current common point cloud semantic segmentation algorithms.

Key words: mandibular angle osteotomy, point cloud segmentation, self-attention, deep learning

中图分类号: