Multi-modal medical image integration is necessary to improve the accuracy of diagnostic results by integrating complementary information between modalities such as Computed Tomography and Magnetic Resonance Imaging. Nevertheless, current methods of fusion fail to maintain the finer details of the anatomy, balance of contrast and eliminate artifacts along structural boundaries. In order to solve these issues, this paper will present a new fusion framework that is based on Covariance Filtering. It uses a two-scale image decomposition to extract the base and detail layers and proceeds with saliency-guided weight map creation and covariance-based refinement for spatial consistency and edge preservation. Assessments of multi-modal brain imaging data show that the suggested method is better than the state-of-the-art methods in terms of Entropy, Mutual Information, Average Gradient and QAB/F. All the research findings support its efficacy in maintaining anatomical boundaries, improving visual clarity, and offering a computationally efficient solution. This framework will be expanded to 3D image fusion in future work and dedicated mechanisms used in regions detection to enhance precision in diagnostic processes.
Sharma et al. (Thu,) studied this question.