文档介绍：摘要
本文的主要工作是关于对医学图像分割算法的研究。通过研究医学图像分割前对图
像的预处理、具体的分割算法、分割后在实际中的应用及最后对分割算法的评价准则，
对医学图像分割算法做了一个比较全面的介绍。重点介绍了对 Live Wire 算法的两种
优化策略。大大提高了医学图像分割的速度、效率和准确性，在实际使用中有着重要
的理论价值和应用价值。
本文首先介绍了一些比较经典的平滑和锐化算法，为后期的分割做预处理；接着介
绍了三类经典的医学图像分割算法；通过研究 Live Wire 算法的特点，在 Dijkstra 算
法中应用堆排序，使算法的时间复杂度降为O(nlog n) ；在该算法的最短路径搜索中加
入到达目标节点即停止的限制条件，使算法的时间复杂度远远小于O(nlog n) ，有效提
高 Live Wire 算法运行的效率。
最后介绍了医学图像分割在二维影像工作站及三维医学图像可视化中的应用。

关键词：图像预处理医学图像分割图像分割
ABSTRACT
This article’s prime task is about the Medicine Image Segmentation
study Medicine Image Segmentation after image pretreatment,segmentation application in
reality and to segmentes the algorithm finally the appraisal criterion, it has made a quite
comprehensive introduction to the medicine image segmentation algorithm. Introduced with
emphasis to Live the Wire algorithm two kind of optimized strategy. Raised the medicine
image segmentation speed, the efficiency and the accuracy greatly, has the important theory
value and the application value in the actual use.
This article first introduced some quite classical smooth and peaked algorithm, makes
the pretreatment for later period's segmentation; Then introduced three kind of classical
medicine image segmentation algorithm; The fourth chapter mainly through studies the Live
Wire algorithm the characteristic, through applies piles sorting in the Dijkstra algorithm,
causes the algorithm the time order of complexity to reduce to O(nlog n ) ; Joins in this
algorithm's most short-path search arrives at the goal node is the stop limiting condition,
causes the algorithm the time order of complexity is smaller than O(nlog n ) by far, by
enhances Live effectively the Wire algorithm movement the efficiency.
At last, we introduced medicine image segmentation application finally in reality; And
de