Multimodal Image Reconstruction Using Supplementary Structural Information in Total Variation Regularization

摘要

In this paper, we propose an iterative reconstruction algorithm whichrnuses available information from one dataset collected using one modality to increasernthe resolution and signal-to-noise ratio of one collected by another modality. Thernmethod operates on the structural information only which increases its suitabilityrnacross various applications. Consequently, the main aim of this method is to exploitrnavailable supplementary data within the regularization framework. The source ofrnprimary and supplementary datasets can be acquired using complementary imagingrnmodes where different types of information are obtained (e.g. in medical imaging:rnanatomical and functional). It is shown by extracting structural information from thernsupplementary image (direction of level sets) one can enhance the resolution of thernother image. Notably, the method enhances edges that are common to both imagesrnwhile not suppressing features that show high contrast in the primary image alone.rnIn our iterative algorithm we use available structural information within a modifiedrntotal variation penalty term. We provide numerical experiments to show thernadvantages and feasibility of the proposed technique in comparison to otherrnmethods.