首页|一种骨科手术导航系统中的2D/3D医学图像配准精度验证新方法

一种骨科手术导航系统中的2D/3D医学图像配准精度验证新方法

扫码查看
目的:2D/3D医学图像配准是骨科手术导航系统中的核心技术,为了定量评估2D/3D图像配准的准确性,文章提出了一种新的图像配准精度验证方法.方法:利用自动配准算法,对每组X射线和CT图像进行配准获得数据作为正样本,并在正样本的基础上增加随机偏移量作为负样本,然后按照新方法和重投影误差对正负样本进行精度验证.结果:新方法验证的正负样本的配准误差为(0.19±0.044)mm,(10.28±3.07)mm.重投影误差验证正负样本中的配准误差分别为(0.304±0.073)mm、(27.816±9.506)mm,图像配准效果及配准误差和配准矩阵之间关系表明本文提出的方法与配准效果正相关,能够定量评估配准结果的准确性.结论:文章提出的方法对2D/3D配准精度的评估优于重投影误差,在排除识别误差和标定误差的情况下,能够定量评估CT图像和X射线图像配准的准确程度,为骨科手术导航系统的应用提供可靠的技术支持.
Validation of 2D/3D Medical Image Registration Accuracy for Orthopedic Surgery Navigation Systems
Objective:The registration of 2D/3D medical images is a crucial aspect of orthopedic surgical navigation systems.This paper presents a novel approach to quantitatively evaluate the precision of image registration and validate its accuracy.Methods:A GPU-accelerated automatic registration algorithm for X-ray and CT images was used to obtain positive samples and a random offset was added to the positive sample results as the negative samples.Then the accuracy was verified according to the new method group and the reprojection error group.Results:The registration error of the new method group verification for positive and negative samples were(0.304±0.073)mm,(27.816±9.506)mm.Reprojection error sets verify that the registration error sets in the positive and negative samples were(0.304±0.073)mm,(27.816±9.506)mm.The image registration effect and the relationship between the registration error and the registration matrix showed that the method proposed in this paper was positively correlated with the registration effect and can quantitatively evaluate the accuracy of the registration results.Conclusion:The presented technique showcasing the registration accuracy was better than the reprojection error.Disregarding recognition and calibration errors,it enables a numerical assessment of the accuracy of CT and X-ray image registration.This offers dependable technical assistance for the use of orthopedic surgical navigation systems.

orthopaedicsnavigationimage registrationaccuracy verification

陈维、何滨、徐琦、陈落根、王沙沙、张巍

展开 >

浙江省医疗器械检验研究院 (浙江 杭州 310018)

杭州三坛医疗科技有限公司 (浙江 杭州 310030)

浙江大学软件学院 (浙江 宁波 315000)

骨科 导航 图像配准 精度验证

2024

中国医疗器械信息
中国医疗器械行业协会

中国医疗器械信息

影响因子:0.375
ISSN:1006-6586
年,卷(期):2024.30(15)