In medical diagnostics, multimodal using complementary data from many imaging modalities, Fusion of medical images has become a crucial tool for enhancing clinical decision-making. This research work presents a novel paradigm for fusion framework that integrates salient structure extraction methods with bilateral and Gaussian filters. The suggested method efficiently synthesizes structural details and prominent characteristics from source images, utilizing complex algorithms like the Swift transformation to enhance and extract essential information. The use of a decision map guarantees appropriate feature selection, while normalized convolution and linear combination processes facilitate the seamless integration of input images. The resultant merged images provide enhanced clarity and detail, facilitating improved visualization and diagnosis. The implementation of the proposed approach on MRI and PET images illustrates its efficacy, showcasing improved performance in preserving essential characteristics from both modalities. This system provides a dependable and efficient solution for clinical imaging activities, enhancing medical image analysis.