基于小波变换的多方向加权聚类颅内肿瘤图像分割方法

摘要

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 high-frequency 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.%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。