2-Step Sparse-View CT Reconstruction with a Domain-Specific Perceptual Network
This work provides a method to improve the quality of CT reconstructions from sparse-view measurements, which is beneficial for medical imaging and industrial inspection where dense sampling is often impractical or harmful.
The paper addresses the challenge of sparse-view computed tomography (CT) reconstruction, which typically suffers from streak artifacts due to angular undersampling. They propose a two-step framework that first uses a super-resolution network (SIN) on sparse projections to reduce ill-posedness, followed by a refinement network (PRN) on the reconstructions, achieving a 4 dB improvement over current solutions.
Computed tomography is widely used to examine internal structures in a non-destructive manner. To obtain high-quality reconstructions, one typically has to acquire a densely sampled trajectory to avoid angular undersampling. However, many scenarios require a sparse-view measurement leading to streak-artifacts if unaccounted for. Current methods do not make full use of the domain-specific information, and hence fail to provide reliable reconstructions for highly undersampled data. We present a novel framework for sparse-view tomography by decoupling the reconstruction into two steps: First, we overcome its ill-posedness using a super-resolution network, SIN, trained on the sparse projections. The intermediate result allows for a closed-form tomographic reconstruction with preserved details and highly reduced streak-artifacts. Second, a refinement network, PRN, trained on the reconstructions reduces any remaining artifacts. We further propose a light-weight variant of the perceptual-loss that enhances domain-specific information, boosting restoration accuracy. Our experiments demonstrate an improvement over current solutions by 4 dB.