摘要
影像融合是將兩張不同的來源影像，以融合方法將影像合而為一的過程。融合後的影像將會比原始影像包含更多的資訊。醫學影像融合後，將包含更多的資訊在融合影像中。有了這些融合後的影像，讓醫師在進行病灶診斷時，可以有效的提升診斷的準確度。CT醫學影像提供較明顯的硬組織結構，例如顱骨。然而，MRI影像則是提供軟組織的成像，例如腦部組織。在本文中，我們將使用CLAHE方法來增強影像的對比，再將對比增強後的影像，最後以小波方法進行影像融合。我們使用了7種不同的小波方法來融合CT以及MRI醫學影像。融合後的影像將會包含CT的結構性資訊，以及MRI的功能性資訊，使醫生可以依據此融合影像做更精確的診斷。為了驗證本文方法的有效性，我們使用資訊熵(Entropy)、標準差(Standard Deviation)，以及平均梯度(Meangradient)來作為影像評估的指標。實驗的結果顯示，經過本文方法處理的影像，其影像的資訊含量皆高於來源CT影像以及MRI影像。
並列摘要
Image fusion means the process of combining relevant information from two or more images into a single image. The resulting image is more informative than any of original images. Moreover, image fusion on medicine provides more information and it leads to important clinical significance and efficacy for doctors' precise diagnosis. CT medical images provide more obvious hard tissue structure, such as skulls; whereas, MRI image is to provide move visible soft tissue structure, such as brain tissues. In this paper, image fusion of CT & MRI had been carried out by using Contrast Limited Adaptive Histogram Equalization (CLAHE) and wavelet methods. The medical images of CT and MRI were fused by using seven different wavelet transform methods- Boir, coif, db, dmey, haar, rbio and sym. The fused image contained both structural features and functional features; therefore, doctors can make a better clinical diagnosis and decision based on the fused images. In order to verify the effectiveness of this method, we used measuring parameters, such as entropy, standard deviation and mean-gradient to evaluate images. Experimental results showed that the fused images processed by our methods contain more information than any of original CT and MRI images.