CT reconstruction topics

Each topic page explains the problem, picks a few papers to start with, and lists the papers, datasets and toolkits filed under it.

• Surveys & Reviews (25) — Overview papers, tutorials and comparative studies.
• Classical & Iterative Reconstruction (98) — Analytical (FBP/FDK) and iterative (SART/SIRT/CGLS) reconstruction, hand-crafted regularizers (TV, TGV, non-local).
• Sparse-View CT (202) — Reconstruction from a reduced number of projection angles.
• Limited-Angle Tomography (65) — Missing angular wedge, where the null space makes the problem severely ill-posed.
• Low-Dose CT Denoising (126) — Noise suppression at reduced tube current, in image, projection or dual domain.
• Metal Artifact Reduction (59) — Beam hardening, photon starvation and scatter caused by high-density implants.
• Spectral & Dual-Energy CT (42) — Multi-energy acquisition, material decomposition, photon-counting CT.
• Cone-Beam CT (102) — Circular, helical and non-standard CBCT geometries; on-board imaging.
• Dynamic & 4D CT (31) — Time-resolved reconstruction, motion modelling, perfusion and cardiac CT.
• Scatter Correction (12) — Scatter estimation and removal, both hardware- and software-based.
• Geometry Calibration & Motion Compensation (29) — Geometric self-calibration, marker-free alignment, rigid and non-rigid motion correction.
• Deep Unrolling & Model-Based Learning (75) — Learned iterative schemes that unroll a classical solver, with or without data consistency.
• Self-Supervised & Untrained Methods (51) — Training without paired data — deep image prior, noise2noise-style, unsupervised regularization.
• Diffusion & Generative Priors (73) — Score-based, diffusion and other learned generative priors used as priors or samplers for tomographic inversion.
• Implicit Neural Representations (31) — Coordinate-based networks (INR / NeRF) that represent the volume itself as the unknown.
• Gaussian Splatting (29) — Explicit Gaussian primitives as a differentiable tomographic representation.