一种小鼠CT图像分割方法

Organ Segmentation for Mouse CT Image

由于micro-CT图像的成像原理和组织特性导致的多样性和复杂性,针对小鼠CT图像,单一的图像分割方法难以取得良好的分割结果。结合当前医学图像分割方法,提出将阈值分割和基于Amira软件的交互式分割相结合的方法。针对不同的器官组织,根据对象灰度特征、形状、空间分布以及边缘分布等信息,使用不同的分割方法及工具进行逐一分割,最后将不同的器官融合成完整的分割后CT图像。该半自动交互式分割方法结合了两者的特点,可实现小鼠CT图像的准确分割。

a single image segmentation method is difficult to obtain good segmentation results because of the complexity of imaging principle and the diversity caused by tissue specificity. We need to design a proper segmentation scheme for Micro- CT images. By investigating the current medical image segmentation methods, we propose a method combining the auto threshold segmentation and Amira software-based interactive segmentation. Specifically, according to the object gray fea- ture, and shape, spatial distribution and edge distribution information, different methods and tools are adopted for differ- ent organs, and then, integrate different argans into a complete segmented CT image. Due to combining the advantage of threshold segmentation and Amira software, this semi-auto segmentation method can realize accurate segmentation of mouse CT images and avoid the high cost for improving the image quality, which also lay a good foundation for the subse- quent reconstruction.