一种基于数据融合的医学图像分割方法

Medical Image Segmentation Method Based on Information Fusion

针对一类纹理特征明显的医学图像 ,提出了一种融合纹理信息和灰度信息的图像分割方法 ,设计了基于金字塔结构的区域增长分割算法 .该方法在区域内部结合使用纹理信息和灰度信息 ,在区域边缘部分则充分利用灰度信息 .计算结果表明 ,该方法对某一类医学图像能够获得较好的分割效果 .

A new method for medical image segmentation with adequate texture information is proposed, in which texture and gray variation information were integrated. A pyramid data structure-based region growing algorithm was designed and applied to medical image segmentation. In this algorithm, the combination of texture information and gray variation information was implemented inside the region. Meanwhile, the former was mainly used near the edge of region. The experimental result shows that this method can produce good segmentation result when applied to certain medical images.