Medical Image Processing Group, Institute of Automation, Chinese Academy of Sciences, Beijing P.O. Box 2728, 100080, China.
J Xray Sci Technol. 2003 Jan 1;11(4):193-204.
In this paper, an interactive segmentation method that combines fast marching and level set method is proposed. Level set segmentation involves solving the energy-based active contour minimization problems by the computation of geodesics or minimal distance curves. First, by selecting the seed point, the fast marching method is used to extract rough boundaries of the interested object. We modified the traditional fast marching method to capture the weak edges by introducing watershed transform. Then, the contour obtained from the fast marching method mentioned above is regarded as an initialization and the level set method is used to finely tune the contour. The algorithm is demonstrated on some medical images: segmentation of knee tissues in CT image and segmentation of brain tissues in MR image. The results show that this method can remove the small regions obtained from fast marching method and converge the desired boundary.
本文提出了一种结合快速行进法和水平集方法的交互式分割方法。水平集分割涉及通过计算测地线或最小距离曲线来解决基于能量的主动轮廓最小化问题。首先，通过选择种子点，快速行进法用于提取感兴趣对象的粗略边界。我们通过引入分水岭变换来修改传统的快速行进法以捕获弱边缘。然后，将从上述快速行进法获得的轮廓视为初始化，并使用水平集方法来精细调整轮廓。该算法在一些医学图像上进行了演示：CT 图像中膝关节组织的分割和 MR 图像中脑组织的分割。结果表明，该方法可以去除快速行进法得到的小区域，并收敛到期望的边界。
