Chang Herng-Hua, Valentino Daniel J
Biomedical Engineering IDP, University of California at Los Angeles, Los Angeles, CA 90095-1721, USA.
Comput Med Imaging Graph. 2008 Jan;32(1):22-35. doi: 10.1016/j.compmedimag.2007.08.012. Epub 2007 Oct 15.
A new deformable model, the charged fluid model (CFM), that uses the simulation of charged elements was used to segment medical images. Poisson's equation was used to guide the evolution of the CFM in two steps. In the first step, the elements of the charged fluid were distributed along the propagating interface until electrostatic equilibrium was achieved. In the second step, the propagating front of the charged fluid was deformed in response to the image gradient. The CFM provided sub-pixel precision, required only one parameter setting, and required no prior knowledge of the anatomy of the segmented object. The characteristics of the CFM were compared to existing deformable models using CT and MR images. The results indicate that the CFM is a promising approach for the segmentation of anatomic structures in a wide variety of medical images across different modalities.
一种新的可变形模型——带电流体模型(CFM),它通过对带电元素进行模拟来分割医学图像。泊松方程分两步用于引导CFM的演化。第一步,带电流体的元素沿传播界面分布,直至达到静电平衡。第二步,带电流体的传播前沿根据图像梯度发生变形。CFM具有亚像素精度,仅需设置一个参数,且无需分割对象解剖结构的先验知识。使用CT和MR图像将CFM的特性与现有的可变形模型进行了比较。结果表明,CFM是一种很有前景的方法,可用于分割不同模态的各种医学图像中的解剖结构。