基于改进GFF算法的CT与MRI图像融合  

A Novel Method for CT and MR Image Fusion Based on Guided Filter Fusion Technique

作　　者：周啸虎[1] 薛海林[1] 陈广浩[1] 高伟[1] 

机构地区：[1]南京医科大学附属南京医院(南京市第一医院)放射科,江苏省南京市长乐路68号210006

出　　处：《中国数字医学》2017年第12期37-40,共4页China Digital Medicine

摘　　要：目的:提出一种改进的引导滤波融合算法,并将其应用于CT与MR图像融合。方法:首先采用基于线性最小平方误差准则的滤波器获得多尺度图像,然后对不同层次加权图像进行重建,最后采用引导滤波器和加权平均法对主体和细节图像进行融合。结果:4组医学图像融合实验定性与定量分析表明,基于本文提出算法获得的融合图像视觉效果更佳,且定量评价指标互信息测度、解剖相似性测度、边缘信息量测度、相位一致性测度、局部质量指数和视觉信息保真度均优于传统的GFF算法。结论:基于提出的算法适合含噪声的图像融合,具有较高普适性和实用性,是一种可行、稳健的医学图像融合算法。Objective： We propose a novel method for the fusion of CT and MR image fusion based on guided filter fusion （GFF）. Methods： The main steps of the GFF method are filtering （to obtain the two-scale representation）, weight maps construction, and fusion of base and detail layers （using guided filtering and weighted average method）. Linear minimum mean square error estimator （LMMSE） based filter reduces the image noise and the more levels of weight maps ensure that more information is transferred to the fused image. Results： Simulation results based on visual and quantitative analysis show the proposed method has good visual representation, higher definition, better quality and contrast of fused image. Six lneasures are all higher than existing guided filtering scheme. Conclusion： The proposed method can obtain more efficient and accurate fusions results even in the noisy image. It is a feasible image fusion algorithm, and can provide better robustness, superiority performance.

关 键 词：图像融合 引导滤波融合 线性最小平方误差 加权图谱 

分 类 号：R319[医药卫生—基础医学] R445