脑切片图像三维重建中的层间配准算法

Registration Algorithm of the Brain Slice Image 3D Reconstruction

在脑切片图像三维重建过程中,为了解决二维医学图像的层间配准问题,提出一种新的模板配准的算法.该算法根据序列图像的共同模板信息,在各图像中搜索匹配图,再将切片图像移动到理想对齐位置,经动态校正后,以目标质心为矩形左上角,截取图像用以重建。实验结果对比分析表明,本算法能精确处理数字人采集的原始数据,并能为后续的二次数据库开发、图像分割、重建和数字解剖等提供应用支持,优于同类算法。

In order to solve the problem of two-dimensional medical image registration between layers in the process of three-dimensional reconstruction of the brain slice image,a new template registration algorithm was proposed.The algorithm based on a common template of the image sequence,search for a matching diagram in each image,move the slice images to the ideal alignment position,after dynamic correction.Intercept the images for the reconstruction begin from the cancroids of the target rectangle.Experimental results contrast analysis showed that this algorithm can accurately handle the raw collection data of the digital human and is expected to provide application support for secondary development of database,image segment,reconstruction and digital anatomy