基于改进GVF-测地线模型的颈动脉斑块的分割算法  

The Improved GVF-GAC based Carotid Plaque Segmentation Algorithm

作　　者：吴学咏[1] 陶进绪[1] 连娟 贾浩[1] 徐明才[1] 李昕[1] 

机构地区：[1]中国科学技术大学电子工程与信息科学系,合肥230027 [2]解放军105中心医院超声科,合肥230031

出　　处：《生物医学工程研究》2010年第2期101-105,共5页Journal Of Biomedical Engineering Research

摘　　要：为了在超声波图像中提取颈动脉斑块边缘,我们提出了一种基于GVF-测地线模型的图像分割算法。由于超声波图像含有大量噪声,首先使用加权均值空间平滑滤波器对图像进行预处理;再人工画出初始轮廓,分别采用GVF-Snake模型、GVF-测地线模型和改进后的GVF-测地线模型对图像进行分割,比较其结果。实验结果表明,改进后的方法分割精度很高,能够在颈动脉斑块边缘提取中取得非常好的效果。A GVF-GAC based image segmentation algorithm for edge extraction of Carotid Plaque on ultrasound images was proposed.Because there are lots of noise existed in ultrasound images,images were pretreated by using a weighted mean filter at first;then initial contour was drawn manually and GVF-Snake model,GVF-GAC model and improved GVF-GAC model were used to segment the images,the results finally were compared.The results of experiment show that improved GVF-GAC model has a high segmentation-accuracy and can make a very good effect in edge extraction of Carotid Plaque.

关 键 词：图像分割 颈动脉斑块 GVF-Snake模型 GVF-测地线模型 测地线模型 

分 类 号：R318[医药卫生—生物医学工程]