Maximizing Information of Multimodality Brain Image Fusion Using Curvelet Transform with Genetic Algorithm

The existing medical image fusion techniques lack of the ability to produce fused image that can maintain fine details of information content from the source images. In this paper, we introduce curve let transform and Genetic Algorithm (GA). The curve let transform performs better than wavelet transform in preserving visual image content particularly the edges. The use of GA can further refine the features of the fused image, and solve the existing uncertainty and ambiguity in the smooth region of the input image. Our method is beneficial to image fusion techniques whose applications rely on the source information of local images. Our experimental results indicate that our method performs betters than baseline methods in terms of quantitative image fusion performance.