HTC 2022: 2D CT Reconstruction Sinograms for Limited Angle Tomography

Five 2D tomographic training phantoms with full angular data, designed for the Helsinki Tomography Challenge 2022. The dataset contains pre-processed sinograms and metadata for plastic phantoms with a 7 cm diameter, created by Alexander Meaney at the University of Helsinki. It includes example reconstructions and segmentations to facilitate algorithm development for limited-angle computed tomography.

Use Cases

• Developing limited-angle tomography reconstruction algorithms based on subsampled sinograms.
• Benchmarking CT reconstruction techniques against provided ground truth from full-data sinograms.
• Exploring segmentation methods for tomographic images based on the provided example procedure.
• Studying the impact of reduced view angles and complex phantom shapes on reconstruction quality.

Strengths

• Data is pre-processed with background, flat-field, and center-of-rotation corrections, and log-transforms are computed.
• Includes five distinct training phantoms (four with holes, one solid disc) for algorithm development.
• Provides example reconstructions and segmentations to demonstrate desired output orientation and format.

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Helsinki

Collection Method

Measurements from a cone-beam computed tomography scanner of custom plastic phantoms.