基于掩膜优化的多模态医学图像刚性配准

本文基于形态学开运算初步去除图像中无需关注的细小区域和部分扫描床，使用最大类间方差法（Otsu）突出感兴趣的图像部位，Canny算子用于提取富含信息区域的边界信息。使用像素填充技术得到图像配准需要的掩膜，并采用OpenMP并行技术加速掩膜计算过程。最终在配准过程中将掩膜作用于参考图像或浮动图像。测试了多组不...

This paper preliminarily removes irrelevant tiny regions and partial scanning bed from images using morphological opening operation. The Otsu's thresholding method is employed to highlight the regions of interest in the images, and the Canny operator is utilized to extract boundary information from information-rich regions. Pixel filling techniques are adopted to generate the masks required for image registration, and OpenMP parallel computing is implemented to accelerate the mask calculation process. Finally, the masks are applied to either the reference images or floating images during the registration procedure. Multiple sets of tests on...

搜集汇总

背景与挑战

背景概述

该数据集提出了一种基于掩膜优化的多模态医学图像刚性配准方法，通过形态学开运算、Otsu法和Canny算子自动生成掩膜，以去除图像背景和扫描床，并利用OpenMP并行技术加速处理。实验表明，该方法在保证配准质量的同时，能节约一半以上时间，1分钟内可完成约100张图像配准，并已成功集成到治疗计划系统DeepPlan中，具有较高的临床实用价值。