一种基于SUV值和活动轮廓模型的肺癌识别方法

A Method of Segmenting Lung Tumor Based on SUV and Active Shape Model

提出一种基于SUV（standard uptake values）和梯度向量流活动轮廓模型（GVF snake）的半自动识别PET（positron emission tomography）图像肺部肿瘤的方法。该方法结合了SUV值和活动轮廓模型提取病灶区域,提高了分割速度和精度。最后应用该方法对PET肺部区域进行了分割实验,实验结果表明该方法能准确分割肿瘤。

In this paper, a semiautomatic method based on SUV and GVF snake for lung tumor PET image segmentation is presented. In order to improve the segmentation speed and precision, the method combines standard uptake values and active shape model. Experimental results show that the proposed method can exactly segment the tumor.