Shift-invariant discrete wavelet transform-based sparse fusion of medical images

In this work, the idea of using shift-invariant discrete wavelet transform and sparse fusion-based magnetic resonance and computed tomography image fusion technique is presented. Source images from different modalities are split into different scale components along with high-level components using shift-invariant discrete wavelet transform. Approximation components are fused using sparse fusion. Different computed weights are joined to be used with source images to get the required fused image as output. Metrics-based visual and quantitative results clearly indicate the worth of proposed new approach in comparison with other existing fusion strategies.