针对传统配准算法无法适用于成像模糊、对比度低的 X光医学图像的问题,本文提出一种基于轮廓点相似性测度的配准技术.首先引入分块双阈值增强策略来提取 DRR 图像和 X 光图像的边缘轮廓信息;其次,采用高斯加权欧氏距离计算图像轮廓的相似度;最后通过平衡优化器算法进行迭代优化,得到最优的位姿参数.实验结果表明:本文算法能够精确提取模糊 X光图像的边缘轮廓信息,而且可以准确评估其与 CT数据的相似度,平均配准成功率超过 94%,算法效率和鲁棒性优于传统算法,可用于医疗诊断、放射疗法、图像引导手术等医学活动.
Aiming at the problem that the traditional registration algorithm cannot be applied to X-ray medical images with blurred imaging and low contrast,this paper proposes a registration technology based on the similarity measurement of contour points.Firstly,a block-by-block dual-threshold enhancement strategy is introduced to extract the edge contour information of DRR images and X-ray images.Secondly,the Gaussian weighted Euclidean distance of contour points is used to calculate the similarity of the two images.Finally,the optimal pose parameters are obtained by iterative optimization using the balanced optimizer algorithm.The experimental results show that the algorithm in this paper can accurately extract the edge contour information of fuzzy X-ray images,and can accurately evaluate its similarity with CT data,which have average registration success rate exceeding 94%.The efficiency and robustness of this algorithm are better than traditional method.It can be used in medical activities such as medical diagnosis,radiation therapy,and image-guided surgery.