Name
Abstract | ||
|---|---|---|
Iterative image reconstruction for positron emission tomography (PET) can improve image quality by using spatial regularization that penalizes image intensity difference between neighboring pixels. The most 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. |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1109/ISBI.2011.5872687 | ISBI |
Keywords | Field | DocType |
optimisation,image quality,optimization transfer algorithm,quadratic penalties,image intensity,edge-preserving regularization,patch-based regularization,pet,image reconstruction,piece-wise constant blocky artifacts,positron emission tomography,iterative pet image reconstruction,iterative methods,medical image processing,robustness,optimization,optimization problem,penalty function,pixel | Iterative reconstruction,Computer vision,Iterative method,Computer science,Image quality,Robustness (computer science),Regularization (mathematics),Artificial intelligence,Pixel,Regularization perspectives on support vector machines,Penalty method | Conference |
ISSN | ISBN | Citations |
1945-7928 E-ISBN : 978-1-4244-4128-0 | 978-1-4244-4128-0 | 4 |
PageRank | References | Authors |
0.46 | 3 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Guobao Wang | 1 | 86 | 12.68 |
| Jinyi Qi | 2 | 284 | 35.82 |