Limited-Angle Tomography
Limited-angle tomography reconstructs an image when the projections cover only part of the required angular range, leaving a missing wedge of directions. Unlike sparse-view CT, the missing data are contiguous: edges whose orientation falls inside the wedge are essentially invisible to the measurements, so the problem is severely ill-posed and filtered back-projection produces directional streaks and smeared boundaries.
It arises whenever the source and detector cannot rotate fully around the object: C-arm and interventional imaging, digital breast tomosynthesis, laminography of flat parts, in-line industrial inspection and electron tomography.
Approaches include anisotropic and directional total variation and other sparsity priors, hybrid methods that learn only the invisible part, sinogram inpainting networks, dual-domain deep networks and diffusion models used as image priors with model-based data consistency. Because so much is unmeasured, accuracy claims are worth checking on measured data such as the Helsinki Tomography Challenge.
Browse all 46 Limited-Angle Tomography papers in the searchable directory
Start here
If you are new to this topic, these are good first reads, chosen by the maintainers.
- A limited-angle CT reconstruction method based on anisotropic TV minimization (PMB, 2013)
Anisotropic TV that adapts the smoothing to the measured angular range; a clean classical baseline. - Learning the invisible: a hybrid deep learning-shearlet framework for limited angle computed tomography (Inverse Problems, 2019)
Learns only the part of the image that the data provably cannot determine and leaves the rest to sparse regularisation; a principled hybrid. - Deep Learning Computed Tomography: Learning Projection-Domain Weights From Image Domain in Limited Angle Problems (TMI, 2018)
Maps FBP onto a neural network and learns projection-domain weights for limited-angle data. - Dolce: A Model-Based Probabilistic Diffusion Framework for Limited-Angle CT Reconstruction (ICCV, 2023)
A conditional diffusion prior with model-based data consistency, evaluated on real limited-angle data.
Papers
The 30 most recent of 46 papers, newest first.
Browse all 46 Limited-Angle Tomography papers in the searchable directory
Datasets, benchmarks and toolkits
Entries from the directory whose tasks or tags match this topic.
- 2DeteCT (official page) · Dataset — Projection + image. 5,000 real experimental 2D slices of mixed objects, 3 dose levels, multiple source-detector geometries. Excellent for reproducible real-data CT benchmarking.
- Helsinki Tomography Challenge 2022 (official page) · Dataset — Projection only. Real limited-angle sinograms of phantoms and of a lotus root; the accompanying challenge defines the standard limited-angle protocol.
- FleX-ray Walnut CBCT Collection (CWI) (official page) · Dataset — Projection + image. 42 walnuts scanned on three different orbits each, so that artefact-free reference volumes can be computed from the combined views and used as supervised-learning ground truth.
- Carved cheese (FIPS) (official page) · Dataset — Projection only. A single 2D slice of a carved cheese from the Finnish Inverse Problems Society, published at three resolutions together with the measurement matrices that encode the X-ray transform as an explicit linear operator.
- Helsinki Tomography Challenge (official page) · Benchmark — Real measured limited-angle sinograms. A purely learned network trained on synthetic data won; the only model-based entry (CIL, directional TV) placed third — evidence that classical methods stay competitive at limited angle.
- DBT-Reconstruction · Toolkit — Open-source reconstruction toolbox for digital breast tomosynthesis, covering DBT-specific geometry alongside FBP and iterative solvers.
Guides
Guides that cover this topic.
Other topics
- Surveys & Reviews (24)
- Classical & Iterative Reconstruction (71)
- Sparse-View CT (157)
- Low-Dose CT Denoising (69)
- Metal Artifact Reduction (23)
- Spectral & Dual-Energy CT (19)
- Cone-Beam CT (77)
- Dynamic & 4D CT (24)
- Scatter Correction (11)
- Geometry Calibration & Motion Compensation (28)
- Deep Unrolling & Model-Based Learning (68)
- Self-Supervised & Untrained Methods (35)
- Diffusion & Generative Priors (68)
- Implicit Neural Representations (25)
- Gaussian Splatting (26)