压缩感知及其在医学图像融合中的应用

Compressed sensing and its application in medical image fusion

压缩感知在采样的同时完成信号的压缩,为解决医学图像融合过程中时间复杂度高、传输数据量大的问题提供了新方法。本文围绕压缩感知在医学图像像素级融合做了5个方面的工作:第一,给出了基于压缩感知的医学图像融合框架;第二,讨论了基于贝叶斯、贪婪迭代、凸松弛等四类重构算法;第三,梳理出医学图像像素级融合的七类方法;第四总结出基于压缩感知的四种医学图像融合路径;第五,指出了目前研究的难点和应用前景。

Compressed sensing （CS） is a new compression sampling technology, which can reduce the sampling data, storage and transmission. CS provides a new way to resolve high time complexity, large transmission data in medical image fusion. Five aspects of pixel-level fusion in medical image are discussed in this paper. Firstly, a framework of medical image fusion based on com- pressed sensing is putted forward by this paper. Secondly, four kinds of reconstruction algorithm, such as Bayesian, convex relaxa- tion, greedy iterative, are discussed comprehensively. Thirdly, seven kinds of medical image fusion in pixel-level fusion are sum- marized, Fourth, four paths of medical image fusion based on Compressed sensing are summarized. Finally, The difficulties and application prospects of the research are pointed out.