Loading...

Development of Software Based on Statistical Iterative Algorithm in the Reconstruction of Sparse View CT images

Jamaati, Sayna | 2023

75 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56354 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Hosseini, Abolfazl
  7. Abstract:
  8. X-ray Computed Tomography (CT) is a widely used medical imaging technique that provides cross-sectional images by measuring the attenuation of X-rays in the body. However, the increasing use of CT has raised concerns about the potential long-term risks associated with radiation exposure. Various approaches have been proposed to reduce radiation dose, and one of the latest methods is sparse-view CT, which has gained popularity due to its lower challenges compared to other techniques. In sparse-view CT, data acquisition is restricted to specific angles, resulting in a significant reduction in radiation dose. However, this approach can introduce streak artifacts in the reconstructed images due to the limited data available. To address these artifacts, image reconstruction algorithms play a crucial role. While traditional Filtered Back Projection (FBP) algorithms have shown limitations in sparse-view image reconstruction, Iterative Reconstruction (IR) algorithms like Algebraic Reconstruction Technique (ART) have been employed along with optimization methods such as Total Variation (TV) to mitigate artifacts. Although these methods have achieved some success in removing streak artifacts in sparse-view images but when the number of data is significantly reduced, the loss of sharpness and edges of the image is unavoidable in these methods. In this study, we propose a promising method called OSEM-AANLM (Ordered Subset Expectation Maximization - Adaptive Accelerated Non-Local Means) to address streak artifacts. This method combines a Statistical Iterative Reconstruction (SIR) algorithm with an adaptive accelerated optimization method. The results demonstrate the effectiveness of OSEM-AANLM in improving sparse-view images by preserving image details and successfully removing artifacts without compromising image edges or resolution. Intuitive comparisons and evaluation criteria confirm the superiority of this method over other methods. Based on the values obtained from the evaluation criteria, the results of the proposed algorithm have been improved by 16% compared to the previous methods. Additionally, this research aims to simulate the CT imaging system using GATE software to generate desired outputs and data. Moreover, in this thesis, an image reconstruction software that includes multiple image reconstruction algorithms, interpolation methods, post-processing filters and image quality evaluation methods has been developed
  9. Keywords:
  10. Computed Tomography (CT) ; Image Reconstruction ; Artifact Reduction ; Geant4 Application for Tomographic (GATE)Emission ; Reconstruction Software Development ; OSEM-AANLM Algorithm ; Sparse View

 Digital Object List

 Bookmark

No TOC