Artificial Intelligence for Merging Modalities in Medical Imaging: A Fusion-Based Enhancement Approach

Abstract

Modern medical imaging plays a central role in diagnostics, yet each modality carries unique strengths alongside unavoidable limitations. Drawing on insights from Stanford's Mini-Fellowship in Molecular Imaging Techniques, this study introduces FusionScan, a novel framework that unites Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) with an AI-driven enhancement system. The goal is to evaluate whether this combination can achieve greater depth, resolution, and functional detail by harnessing the complementary advantages of both modalities, offering a more comprehensive solution for complex diagnostic needs. The approach involves merging MRI and CT data and applying AI algorithms to refine resolution and increase molecular specificity. Particular emphasis is placed on the challenge of multimodal alignment, with strategies proposed to ensure precise registration between imaging outputs. Publicly available medical imaging datasets will serve as the testing ground for assessing how AI improves reconstruction, resolution, and adaptive adjustment of imaging parameters. An integrated feedback mechanism is also planned, allowing the system to modify scanning parameters in real time based on preliminary AI assessments. This feature is designed to minimize radiation exposure, shorten scan durations, and concentrate detail on regions requiring closer examination. Ultimately, this research envisions the development of a fused, AIenhanced architecture capable of delivering superior diagnostic insight, supported by simulations that illustrate the transformative impact of AI on medical imaging.