融合颜色和强度先验信息的几何形变模型在医学图像分割中的应用

Geometric Active Contour Model with Color and Intensity Priors for Medical Image Segmentation

研究了一种基于融合颜色和强度先验信息的几何形变模型的医学图像分割算法。首先借助遗传算法,从图像的不同颜色空间中得到对应颜色分量的阈值,然后将这些阈值所代表的先验信息融合到形变模型的进化曲线中,从而得到了改进的基于leveI set的几何形变模型。采用临床骨髓细胞和乳腺细胞分别进行了实验,结果表明相对于传统的只利用图像梯度信息的形变模型．算法不仅能得到更精确的结果,而且具有更快的运行速度。

A new algorithm using the geometric active contour model with the fusion of color and intensity priors to segment medical images is presented in this paper. The prior knowledge used here are firstly defined in different color spaces and represented as thresholds searched by the genetic algorithm. Then the prior knowledge is merged into active contour model with its contour evolution by the level set technique. The experiments on clinical marrow images and mammograms have successfully demonstrated its superiority of the proposed algorithm over the existing active contour models which deal with image gradient information.