基于模糊增强的医学超声图像分割

Segmentation of Ultrasound Images with Fuzzy Enhancement

医学超声图像因对比度较低、形成过程产生的独有的斑点噪声等影响了手动与计算机辅助分析的效果,尤其是影响了计算机定量测量的效果。为了改进超声图像的分割效果,首先要增强超声图像的对比度,本文提出了一种模糊增强算法对超声图像进行显著增强,然后利用马尔可夫随机场与最大后验概率理论对增强后的图像进行分割,得到了比较满意的分割效果。

Ultrasound B-scan images often have low contrast and the characteristi c speckle noise created in the forming process. These cause major problems for i mage analysis, both by manual and computer-aided techniques, particularly the co mputation of quantitative measurements. We present a new fuzzy algorithm to enha nce the contrast of ultrasound images, and use markov random field(MRF) and the maximum a posteriori theory to segmentation the result images. Significant impro vement is achieved in tissue contrast and segmentation result.