STEMPO: Dynamic X-Ray Tomography Phantom Data for Reconstruction Method Validation

The STEMPO phantom dataset includes 2D and 3D data from static, continuous 360-degree, and sequential 8x45 projection scans, with downsampling factors of 4, 8, 16, or 32. Created by Tommi Heikkilä of the University of Helsinki, the data is designed to test and validate novel dynamic tomography reconstruction methods. The dataset includes example algorithms for analytic and iterative reconstruction methods and is intended to be appended with new data in the future.

Use Cases

• Validate dynamic tomography reconstruction methods based on the provided static, continuous, and sequential scan data.
• Benchmark iterative reconstruction algorithms against the included example codes for FBP, FDK, PDFP wavelet, and L+S methods.
• Test the impact of different measurement geometries and sampling setups using the provided geometry parameters.

Strengths

• Includes multiple data types: static reference scans, continuous 360-degree dynamic scans, and sequential 8x45 dynamic scans.
• Provides downsampled versions (binning factors 4, 8, 16, 32) for scalability testing.
• Comes with example implementation codes for several well-known analytic and iterative reconstruction methods.
• Includes an approximated ground truth motion file for the dynamic scans.

Limitations

• Row count and total 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

Data collected from a computer-controlled physical phantom (STEMPO) designed for dynamic X-ray tomography.