Measurement and analysis of multi-modal image fusion metrics based on structure awareness using domain transform filtering

Due to varrying imaging principles and interwined complexity of human organ structures, different types of medical images must be combined, as single-modality medical images may only provide limited information. In this paper, a multimodal medical image fusion method that integrates multimodal medical images having low resolution with reduced computational complexity to improve the accuracy of target recognition and for providing a basis for clinical diagnosis. Initially salient structure extraction (SSE) approach, which employ a rolling guidance filter (RGF) over the source images for removing small scale structures while preserving the image textures and thereby recovering the salient edges has been implemented. Subsequently image gradient operator is employed to restores large-scale structures from the filtered images. A DTF (Domain Transfer Filtering) is further used to recover the small-scale details in the neighborhood of large-scale structures of the images. The output of DTF is used as a weighted map that is combined with the source images to recover fusion result by a weighted-sum rule. Image fusion measurement for quality assessment and objective analysis is carried out using various fusion metrics. Experimental result shows that the proposed method can obtain high quantitative and qualitative performance as compared to other state-of-the-art methods and can eventually provide effective reference for doctors to assess patient condition.