基于小波变换的多方向加权聚类颅内肿瘤图像分割方法  

INTRACRANIAL TUMOUR IMAGE SEGMENTATION ALGORITHM WITH MULTIDIRECTIONAL WEIGHTED AGGREGATION BASED ON WAVELET TRANSFORM

作　　者：段娟[1] 景万里 张楠[2] 肖创柏[1] 武博[2] 

机构地区：[1]北京工业大学计算机学院,北京100124 [2]首都医科大学生物医学工程学院,北京100069

出　　处：《计算机应用与软件》2014年第4期185-187,196,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(61272255);北京市教委科技项目(KM201010005011;KM201310025009)

摘　　要：分割MR颅内肿瘤区域后,可对病变部分进行针对性的临床治疗和处理。结合临床实际,提出一种基于小波变换的多方向加权聚类颅内肿瘤图像分割方法。通过小波变换和多方向加权聚类算法的结合,利用非线性的补偿图像轮廓来增强小波分解后的高频信号,并对低频信号做多方向的种子点选取加权聚类算法改进。实验结果表明,该算法在大大减少时间复杂度的基础上,也能有效地提高准确性。Targeted clinical therapy and treatment can be applied in the lesion part after the segmentation is made on MR intracranial tumour area. With clinical practice,we propose a wavelet transform-based intracranial tumour segmentation algorithm with multi-directional weighted aggregation. Through the combination of wavelet transform and multi-directional weighted aggregation algorithm,it enhances the highfrequency signal decomposed with wavelet by using a non-linear compensation image contours,and improves the low-frequency signals by the multi-directional weighted aggregation algorithm with seed point selection. Experimental results show that on the basis of greatly reducing the time complexity,the proposed algorithm can effectively improve the accuracy.

关 键 词：颅内肿瘤 小波变换 加权聚类 多方向 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]