计算机与数字工程2025，Vol.53Issue(3)：894-900,7.DOI:10.3969/j.issn.1672-9722.2025.03.047
基于可变形模板匹配的三维骨骼测量方法
3D Bone Measurement Method Based on Deformable Template Matching
王菲 1崔青山 1徐楷文1
作者信息
- 1. 河海大学物联网工程学院 常州 213000
- 折叠
摘要
Abstract
In computer aided preoperative planning,the correct extraction of bone medical semantics can help doctors accu-rately measure bone parameters,and then judge whether there exist deformities,and determine the degree of deformity.Traditional bone parameter measurement methods can not meet the actual requirements of preoperative planning,which is laborious and individ-ual differences.The difficulty lies in how to extract bone semantics accurately.This paper proposes a 3D bone measurement method based on deformable template matching.The semantic segmentation of bones is based on deformable template matching.The difficul-ty of matching is to calculate the corresponding relationship.The neural network is used to predict the corresponding semantic region between the bone template and the target bone,extract the center point of the semantic region,construct the sparse correspondence of feature points,and fit the bone template to the target bone according to the sparse correspondence,and then the semantic region of the pre-labeled bone template is mapped to the target bone.Finally,the axis is extracted according to the semantic region of the target bone,and the bone parameters are measured.The experimental results show that the accuracy of Dice coefficient of semantic region obtained by this method is 94%.the average error of neck shaft angle is 3.17°and the average error of anteversion angle is 2.62°by this method.This method can meet the actual needs of bone parameter measurement,and effectively improve the efficiency of preoperative diagnosis and planning of bone deformity.关键词
计算机辅助术前规划/三维骨骼测量/医学语义分割/模板匹配/迭代拟合Key words
computer-aided preoperative planning/three-dimensional bone measurement/medical semantic segmentation/template matching/iterative fitting分类
计算机与自动化引用本文复制引用
王菲,崔青山,徐楷文..基于可变形模板匹配的三维骨骼测量方法[J].计算机与数字工程,2025,53(3):894-900,7.
