基于目标监督的先验形状level set模型分割心脏MR图像

Cardiac MRI Segmentation by Using Level Set Method with Priors Shape Information Based on Object Supervision

提出一种基于目标监督的先验形状level set模型分割心脏MR图像的方法。该方法先根据目标灰度的特征生成特征图像,使得特征图像目标区域灰度趋于一致,再在特征图像的基础上利用先验形状信息的水平集模型对目标进行分割。实验结果表明,该方法能够有效地分割心脏MRI图像。

A new method for segmenting the cardiac MRI image is presented,which is based on object supervision level set method with prior shape information. Feature images are produced according to the similarity of the object intensity. They are segmented by using level set method with priors shape information. Results from the experiment show its effectiveness for segmenting the cardiac MRI images.