Patch-based regularization for iterative PET image reconstruction

Guobao Wang, Jinyi Qi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Scopus citations


Iterative image reconstruction for positron emission tomography (PET) can improve image quality by using spatial regularization that penalizes image intensity difference between neighboring pixels. Themost commonly used quadratic penalty often over-smoothes edges and small objects in reconstructed images. Non-quadratic penalties can preserve edges but may introduce piece-wise constant blocky artifacts. The results are also sensitive to the hyper-parameter that controls the shape of the penalty function. This paper presents a robust regularization for iterative image reconstruction by using neighborhood patches instead of individual pixels in formulating the non-quadratic penalties. An optimization transfer algorithm is developed for the corresponding optimization problem. Computer simulations show that the proposed patch-based regularization can achieve better contrast recovery for small objects compared with quadratic regularization, and is more robust to the hyper-parameter than the conventional pixel-based non-quadratic regularization.

Original languageEnglish (US)
Title of host publicationProceedings - International Symposium on Biomedical Imaging
Number of pages4
StatePublished - 2011
Event2011 8th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI'11 - Chicago, IL, United States
Duration: Mar 30 2011Apr 2 2011


Other2011 8th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI'11
Country/TerritoryUnited States
CityChicago, IL


  • edge-preserving regularization
  • image reconstruction
  • PET

ASJC Scopus subject areas

  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Patch-based regularization for iterative PET image reconstruction'. Together they form a unique fingerprint.

Cite this