Regularization techniques in emission tomography

Project Details


In this project, regularization techniques for image reconstruction in emission tomography will be studied, in order to achieve an improved compromise between noise and resolution. Two groups of regularization techniques will be studied:
1. Regularization by selective suppression of typical noise patterns
2. Regularization based on anatomical information obtained from MR (magnetic resonance) images.
In principle, the same regularization techniques can be applied to almost all forms of emission tomography (2D and 3D PET, SPECT with different collimator types). However, there are two techniques that have received less attention in the past, and for which the implementation of regularized reconstruction will be more challenging: these are pin hole SPECT and reconstruction from list-mode acquisition (both PET and SPECT).

Effective start/end date1/01/0331/12/06

Flemish discipline codes

  • Basic sciences


  • Image Reconstruction
  • Medical Imaging
  • Inverse Problems