4. Demonstrations

• 4.1. Setting up a Python workspace
• 4.2. Exploring raw data and making use of the input-output module
  • 4.2.1. Nxs/hdf files
  • 4.2.2. Tiff files
  • 4.2.3. Mrc files
  • 4.2.4. Other file formats
• 4.3. Methods and tools for removing ring artifacts
  • 4.3.1. Improvements
  • 4.3.2. Tools for designing ring removal methods
    • 4.3.2.1. Back-and-forth sorting
    • 4.3.2.2. Separation of frequency components
    • 4.3.2.3. Polynomial fitting along an axis
    • 4.3.2.4. Wavelet decomposition and reconstruction
    • 4.3.2.5. Stripe interpolation
    • 4.3.2.6. Transformation between Cartesian and polar coordinate system
    • 4.3.2.7. Transformation between sinogram space and reconstruction space
• 4.4. Comparison of ring removal methods on challenging sinograms
  • 4.4.1. Same sample-type and slice but different in shape
  • 4.4.2. Partial ring artifacts
  • 4.4.3. All types of ring artifacts
  • 4.4.4. Having valid stripes (not artifacts)
  • 4.4.5. For cone-beam tomography
• 4.5. Complete workflow for processing tomographic data
  • 4.5.1. Assessing raw data
  • 4.5.2. Reconstructing several slices
  • 4.5.3. Finding the center of rotation
  • 4.5.4. Tweaking parameters of preprocessing methods
  • 4.5.5. Choosing a reconstruction method
  • 4.5.6. Performing full reconstruction
  • 4.5.7. Automating the workflow
  • 4.5.8. Downsampling, rescaling, and reslicing reconstructed volume
  • 4.5.9. Common mistakes and useful tips
  • 4.5.10. Data analysis

Examples of how to use the package are under the example folder of Algotom. They cover most of use-cases which users can adapt to process their own data. Examples of how to process speckle-based phase-contrast tomography is at here.

Users can use Algotom to re-process some old data collected at synchrotron facilities suffering from:

• Various types of ring artifacts.

• Cupping artifacts (also known as beam hardening artifacts) which are caused by using: FFT-based reconstruction methods without proper padding; polychromatic X-ray sources; or low-dynamic-range detectors to record high-dynamic-range projection-images.

There are tools and methods users can use to customize their own algorithms:

• Methods to transform images between the polar coordinate system and the Cartesian coordinate system.

• Methods to separate stripe artifacts.

• Methods to transform an image between the reconstruction space and the sinogram space.

Tomographic data for testing or developing methods can be downloaded from Zenodo.org or TomoBank. Methods can also be tested using simulation data as demonstrated here.