Redundant Discrete Wavelet Transform Based Medical Image Fusion

In this work, we propose redundant discrete wavelet transform (RDWT) based fusion for multimodal medical images. The shift invariance nature of RDWT shows its usefulness for fusion. The proposed method uses maximum scheme for fusion of medical images. We have experimented with several sets of medical images and shown results for three sets of medical images. The effectiveness of fusion results has been shown using edge strength, and mutual information fusion metrics. The qualitative and quantitative comparison of the proposed method with spatial domain fusion methods (Linear, Sharp, and principal component analysis (PCA)) and wavelet domain fusion methods (discrete wavelet transform (DWT), lifting wavelet transform (LWT), and multiwavelet transform (MWT)) proves the superiority of the proposed fusion method.