摘要
目的:分析小波局域熵方法对CT、MRI医学影像的融合效果。方法:选取哈佛大学Whole Brain Atlas数据库CT、MRI脑部影像样本20例,评价图像的信息熵、梯度、标准差等融合指标。选取1例典型样本施行像素级融合、小波系数融合、小波局域熵融合,对3种方法进行比较。聚焦小波局域熵方法,对20例样本的融合指标进行分析,验证小波局域熵的融合有效性和局域窗口优选值。结果:小波局域熵方法优于像素级融合或小波系数融合方法。结果表明,小波局域熵窗口宽度W=7融合效果显著,信息熵提升+29.9%,梯度提升+12.3%,标准差较CT基准优化-21.3%、较MRI基准提升+21.6%。结论:小波局域熵融合方法有效性得到验证,W=7可作为该方法的推荐窗口参数。
Ojective To analyze the effffect of the wavelet local entropy fusion method on the fusion of CT and MRI medical images.Methods Twenty sets of brain CT and MRI images are selected from the Harvard University Whole Brain Atlas database,and fusion evaluation is performed based on indices such as image information entropy,gradient,and standard deviation.For one typical sample,three fusion methods,including pixel﹣level fusion,wavelet domain coeffifficient fusion,and wavelet local entropy fusion,are compared.Subsequently,focusing on the wavelet local entropy method,analysis of fusion indices is conducted on 20 samples to verify the fusion effffectiveness of wavelet local entropy and the optimal value of local window.Results The wavelet local entropy method outperforms both pixel﹣level fusion and wavelet coeffifficient fusion.When the wavelet local entropy window width is W=7,the fusion effffect is signifificant that the information entropy increased by﹢29.9%,gradient increased by﹢12.3%,and standard deviation optimized by﹣21.3%compared with the CT benchmark and increased by﹢21.6%compared with the MRI benchmark.Conclusions The effffectiveness of the wavelet local entropy fusion method has been verifified,and W=7 can be used as the recommended window parameter.
作者
赖伟东
王晓君
潘文
钟立强
LAI Weidong;WANG Xiaojun;PAN Wen;ZHONG Liqiang(College of Optoelectronic Information,Zhongshan Torch Polytechnic,Zhongshan,Guangdong 528436,China;Guangzhou Shirui Electronics Technology Co.,Ltd.,Guangzhou,Guangdong 510700,China)
出处
《影像研究与医学应用》
2025年第19期8-11,共4页
Journal of Imaging Research and Medical Applications
基金
2023年度广东省普通高校特色创新类项目(2023KTSCX363)
广东省普通高校创新团队项目(2024KCXTD057)
中山火炬职业技术学院校级产学研项目(2023CXY13)。
关键词
图像融合
小波局域熵
医学影像
像素级融合
小波系数融合
Image fusion
Wavelet local entropy
Medical image
Pixel﹣level fusion
Wavelet domain coeffifficient
