摘要
针对医学图像的特点,提出了一种结合L ive W ire算法和轮廓插值算法的三维医学图像分割算法。与原L ive W ire算法相比,改进算法在不增加算法复杂度的同时,大大提高了图像分割的性能。为了进一步提高图像的准确性,用下列方法弥补了原算法的不足:通过寻求最短搜索路径,另外尽可能地降低搜索的时间复杂度。将改进的算法与余弦轮廓插值技术相结合,用在病灶组织区域形状变化较大的断层图像(关键层)分割中。实验表明算法能快速有效的从序列医学图像中分割出感兴趣的物体,为医学诊断提供了依据。
According to the characteristics of medical images,a combinated algorithm of Live Wire algorithm and contour interpolation algorithm is proposed for 3D medical image segmentation.Compared with the original Live Wire algorithm,the improved algorithm greatly improves the performance of image segmentation without increasing complexity of the algorithm.And to make up the shortages fo the original algorithm,two measures are taken：（1） to find the shortest search path;（2） to reduce the search time complexity as far as possible.The improved algorithm and cosine interpolation contours are combined for the segmentation of cross-sectional image（critical layer） of lesions area where the shape changes evidently.Experiments show that the algorithm can segment the objects of interest from the sequence of medical images quickly and efficiently.
出处
《计算机仿真》
CSCD
北大核心
2010年第12期262-266,共5页
Computer Simulation
关键词
医学图像分割
交互式分割
轮廓插值
Medical image segmentation
Interactive Segmentation
Contour interpolation
