ABSTRACT
This chapter focuses on a unified approach to the collection of problems arising in the fusion of data from imaging sensors by extracting, aligning, and fusing information from multiple heterogeneous imaging modalities simultaneously with the image reconstruction processes through a unified variational formulation. It presents examples of fusion applied to simulated vascular imaging and real vascular imaging with multidetector computed tomography and volume computed tomography. The chapter attempts to combine the reconstruction, enhancement, and alignment components of this common processing structure into a single unified process. Limited information quality from single modality observations often leads to the desire to combine data from multiple, complementary sensors. The hope is that information that is weakly present in each modality or sensor will reinforce each other when combined, thereby rising above the background and yielding more reliable estimates. With unified processing, it is possible to simplify interpretation of fusion results since reconstruction and alignment occur at a common resolution.
