基于多agent系统的医学图像进化分割算法的研究

Evolutional algorithm of medical image segmentation based on a multi-agent system

在进行医学图像分析时,很多研究对象(如大脑、心脏等)在图像中并没有明显的边界,属于自然纹理图像,不同组织间也没有清晰的分界线,在这种情况下,图像分割的任务非常困难.本文基于多Agent思想提出了进化分割算法.该算法将Agent设计为具有4种行为的计算实体,它驻留在医学图像的二维网格中,利用先验知识指导其行为的演化.通过在人脑核磁共振(magnetic resonance,MR)图像分割和异常脑细胞的识别实验,与最大似然(maximum likelihood,ML)分割和构形梯度平方残差(conjugate gradient square,CGS)分割比较,本文的方法更适合临床.

Many research objects （such as human brain, heart and others） have no obvious brim. When we analyse medical image, it is a natural texture image. There is no clear boundary between different organizations. An evolutional algorithm of medicine image segmentation based on the multi-agent system was proposed in this paper. The agent is designed to be distributed calculation entities of the four entities of the calculation of the distribution in algorithm. It exists in the medical image in the two-dimensional grid, using the priori knowledge to guide its evolution. Through MR image segmentation of the human brain and the identification experiment of abnormal brain cells, compared with maximum likelihood segmentation and conjugate gradient square segmentation, our method is more suitable for clinical application.