改进的医学CT图像K-均值聚类分割算法的实现

Implementation of Improved K-means Clustering Segmentation Algorithm for Medical CT Image

为提高K-均值聚类算法在医学CT图像分割上的应用效果、稳定性和质量,减少程序运行时间,本研究用Matlab语言优化了K-均值聚类算法程序,与Statistics Toolbox的K-means函数进行比较,使用单因素方差分析法检验两种算法实现程序运行时间的差异,并直接观察分割效果和稳定性。结果显示,改进后的K-均值聚类算法程序具有分割结果稳定、质量提高等优点,在常用Windows操作系统和PC机配置环境下,分割耗时在1 s左右,显著低于原有的分割程序,消除了等待感觉,提高了使用者的工作舒适度和效率,为图像的识别处理奠定了基础。

In order to improve the application effect, stability and quality of K-means clustering algorithm in medical CT image segmentation and reduce the running time of the program, K-clustering algorithm program was optimized with Matlab language in this study, and the CT image segmentation experiments were also done on the original medical CT images data set with the optimized K-clustering algorithm program and the K-means function of the MATLAB R2012a Statistics Toolbox. It is analyzed the running time differences of the two kinds of program with single factor variance analysis method, and the stability and quality of the images segmentation results were observed. The results show that the improved K-means clustering algorithm programming had higher stability and quality of segmentation results, etc. Under the environment of ordinary Windows operation system and hardware of personal computer configuration, the segmenting time was about one second, significantly lower than that of the original segmentation procedures, thus the feeling of waiting was eliminated, the user＇s comfort and efficiency were improved, and the foundation was laid for the recognition processing of the image.