Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (9): 155-158.DOI: 10.3778/j.issn.1002-8331.2010.09.044

• 图形、图像、模式识别 • Previous Articles     Next Articles

Brain tumor image segmentation method based on Chan-Vese model

XU Cun-lu1,GAO Jia1,WU Guo-de2   

  1. 1.School of Information Science and Engineering,Lanzhou University,Lanzhou 730000,China
    2.Department of Neurology,Second Affiliated Hospital of Lanzhou University,Lanzhou 730000,China
  • Received:2009-04-07 Revised:2009-06-02 Online:2010-03-21 Published:2010-03-21
  • Contact: XU Cun-lu

Chan-Vese模型下的脑肿瘤图像分割方法

许存禄1,高 佳1,武国德2   

  1. 1.兰州大学 信息科学与工程学院,兰州 730000
    2.兰州大学 第二附属医院 神经内科,兰州 730000
  • 通讯作者: 许存禄

Abstract: A brain tumor image segmentation and 3D reconstruction method is proposed based on Chan-Vese model.The brain tumor image is first eroded iteratively by using morphological erosion to extract the brain tumor contour.Then the brain tumor included in eroded image is segmented by the Chan-Vese model.The image is finally dilated iteratively by using morphological dilation to restore the image.Moreover,all segmented brain tumor images are reconstructed and located to three dimensions.Experimental results show that the algorithm based on Chan-Vese model can solve the problem of incomplete segmentation easily occurred in brain tumor segmentation process.At the same time,the 3D reconstruction and location of the brain tumor offer a great reference value in clinical work.

Key words: brain tumor, image segmentation, Chan-Vese model, 3D reconstruction and location

摘要: 提出了一种基于Chan-Vese模型的脑肿瘤图像分割与三维重构方法。该方法首先通过对脑肿瘤图片的迭代腐蚀操作提取脑肿瘤轮廓,然后利用Chan-Vese模型对脑肿瘤进行分割,最后对图像进行迭代膨胀操作复原图像。另外对所有分割后的脑肿瘤图片进行了三维重构与定位。实验结果表明,基于Chan-Vese模型的图像分割方法很好地解决了脑肿瘤分割过程中容易出现的不完全分割问题,同时对脑肿瘤的三维重构与定位也具有较大的临床实用价值。

关键词: 脑肿瘤, 图像分割, Chan-Vese模型, 三维重构与定位

CLC Number: