超声乳腺肿瘤图像的边缘提取

目的探求乳腺肿瘤超声图像的边缘提取。方法广义梯度矢量流Snake模型已经成功地用于噪声相对比较小的CT、MRI等医学图像,然而乳腺肿瘤超声图像对比度低,斑点噪声大,很难将该模型直接应用于乳腺肿瘤超声图像。本文针对乳腺肿瘤超声图像的特点如图像对比度低,斑点噪声大,部分边缘缺失,肿瘤内部微细结构分布复杂(如血管,钙化灶等),特别恶性肿瘤还具有复杂形状等,采用相应的图像处理技术如非线性各向异性扩散滤除斑点噪声,形态学滤波器平滑图像,直方图均衡化提高图像的对比度,最后将该模型引入到乳腺肿瘤超声图像边缘提取。结果实验对158例乳腺肿瘤超声图像进行边缘提取,定量和定性分析均获得满意的结果。结论本文方法可以有效地用于超声乳腺肿瘤图像的边缘提取。

Boundary extraction of ultrasonic breast tumor image

Objective To extract boundary of ultrasonic breast tumor image. Methods Generalized gradient vector flow snake has been successfully applied to CT and MRI. However, due to poor image contrast and high-level speckle noise it is unsuitable to directly apply GGVF Snake to segment ultrasonic breast tumor image. To deal with the characteristics of the images: poor image contrast, high-level speckle noise, boundary gaps and irregular shape, ultrasonic breast tumor image is firstly preprocessed such as nonlinear anisotropic diffusion, morphological filter and histogram equalization, then GGVF Snake is applied to extract the boundary of breast tumor. Results Experiments on 158 ultrasonic breast tumor images showed that both quality and quantity results are promising. Conclusion Proposed method could extract the boundary of breast tumor effectively.