CAJ | 학술논문

为了增加颅脑肿瘤的诊断信息，提出了基于NSCT(Nonsubsampled Contourlet Transform)和改进的人工蜂群算法(Artificial Bee Colony,ABC)的颅脑CT（Computed Tomography）图像和MRI(Magnetic Resonance Im-aging)图像的配准方法。首先将参考图像和待配准图像进行NSCT变换,分解成高频子带和低频子带，分别提取两幅图像的低频图像作为参考图像和待配准图像，以互信息作为相似性测度，选用刚体变换模型求解空间变换参数，然后提出一种改进的人工蜂群算法来优化配准所需的空间变换参数。实验结果表明，该方法可以有效提高配准速度，具有较好的配准效果。
위료증가로뇌종류적진단신식，제출료기우NSCT(Nonsubsampled Contourlet Transform)화개진적인공봉군산법(Artificial Bee Colony,ABC)적로뇌CT（Computed Tomography）도상화MRI(Magnetic Resonance Im-aging)도상적배준방법。수선장삼고도상화대배준도상진행NSCT변환,분해성고빈자대화저빈자대，분별제취량폭도상적저빈도상작위삼고도상화대배준도상，이호신식작위상사성측도，선용강체변환모형구해공간변환삼수，연후제출일충개진적인공봉군산법래우화배준소수적공간변환삼수。실험결과표명，해방법가이유효제고배준속도，구유교호적배준효과。
In order to increase the diagnosis cerebral tumor ,CT and MRI image registration based on the NSCT and artificial bee colony algorithm is proposed. Firstly, the reference image and input image are decomposed into high frequency subband and low frequency subband with NSCT, and low frequency parts of the two images are ex-tracted as reference image and input image. Mutual information is used for similarity measure,and rigid transform model is adopted to solve the space transform parameters. Then put forward an improved artificial bee colony algo-rithm to optimize the space registration transform parameters. The experimental results show that the algorithm can effectively improve the registration speed and has good effect of registration.