一种基于NSCT和自适应PCNN医学图像融合的改进算法

A Medical Image Fusion Improved Algorithm Based on NSCT and Adaptive PCNN

由于医学影像成像原理不同,不同模态图像的质量、空间和时间特性有较大区别,因此临床上需要将不同模态的图像融合进行综合分析。为了准确并全面地融合MRI和PET图像,提出了一种基于NSCT和自适应PCNN医学图像融合的改进算法。算法采用NSCT得到图像的低频和各方向子带信息,采用符合人类视觉系统的自适应PCNN选择融合图像系数,重构得到融合图像。采用颅脑MRI和PET医学图像进行了仿真实验。结果表明,基于NSCT自适应PCNN算法在提高空间纹理细节和减少颜色失真方面明显优于其它的融合算法。

Due to the different medical imaging imaging principles, different multimodal images quality, space and time characteristics are quite different,and there require fuse different multimodal image to comprehensively analysis in clini-cal.In order to accurately and comprehensively fuse multimodal image, this paper propoed a medical image fusion im-proved algorithm based on NSCT and adaptive PCNN.Algorithm used NSCT to acquire image low and each directional subband information,then uses adaptive PCNN in accordance with the human visual system to select image fusion coef-ficient,finally acquired fused image by reconstruction method.Using brain MRI and PET medical image to simulate ex-periments,the result shows that the algorithm of NSCT adaptive PCNN is much better than other fusion algorithms in the aspects of improving the spatial texture detail and reducing color distortion.