右心室磁共振图像分割算法研究

Research on segmentation algorithm of the Right Ventricles in Magnetic Resonance Images

针对右心室心腔几何形状复杂,解剖结构特殊,心脏磁共振图像难以准确分割。为了使右心室磁共振图像的形态结构和功能异常的检测和诊断对心脑血管疾病能起到很好的作用,通过对阈值分割法、区域生长分割法、K-Means聚类分割方法以及基于阈值水平集分割方法对右心室磁共振图像进行分割算法分别做比对,实验结果表明,通过基于阈值水平集的分割方法具有更高的有效性,分割结果较为准确。

It is difficult to accurately divide the magnetic resonance image of the right ventricle with complicated geometry and special anatomical structure. In order to make the right ventricular morphology and function of the magnetic resonance imaging detection and diagnosis of cardiovascular and cerebrovascular diseases have done a good job, based on threshold segmentation method, region growing segmentation method, and K - Means clustering segmentation method based on threshold level set method of right ventricle mri image segmentation algorithms are checked respectively, the experimental results show that based on threshold level set segmentation method has higher efficiency, more accurate segmentation results.