The effect of sampling in histogramming and analytical reconstruction of 3d AX-PET data
Akram, Mohammad Minhaz (2012)
2012
Master's Degree Programme in Information Technology
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2012-01-11
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201203121055
https://urn.fi/URN:NBN:fi:tty-201203121055
Tiivistelmä
Improvement in both sensitivity and resolution is achieved by the novel AX-PET concept. This design incorporates axial orientation of long crystal (for coincident detection radially) and orthogonally placed wave length shifter strips (for axial detection). Iterative reconstruction of AX-PET list-mode data was performed by the AX-PET collaboration. But this reconstruction takes long processing time due to large dataset and complexity of algorithm. On the contrary, 3D Reprojection (3DRP) is an analytical image reconstruction technique which can be used to reconstruct the 3D list-mode data of two modules AX-PET demonstrator in shorter time.
In this study, 3DRP method is used for image reconstruction. This process requires histogramming and histogramming requires sampling of the coincident events. Obtaining optimum sampling for better resolution is one of the main goals in the image processing. PET scanner acquires data both in axial and radial direction. Hence, optimum sampling rates for axial, radial and angular directions are required. In this study, histogramming is done with different sampling rates to see their effect in the resolution of the reconstructed images.
The construction of the scanner causes gaps and non-homogeneously sampled Filled of View (FOV). Gap filling method fill the gaps generated from inter crystal and inter module gaps. Coincident detection differs due to difference of the positions of crystals. Normalization is required to have the uniform response from all the detectors. In this study normalization and gap filling were used before 3DRP reconstruction.
Preliminary result shows the ability of resolving the capillary of 1.4 mm. But for the complex phantoms the smallest resolvable inserts was 2 mm. The best results were obtained with the sampling rate of 1mm compared to other sampling rates for different phantoms due to better statistics. The list-mode data of the phantoms used in this study is a low statistics data. To exploit the novelty of AX-PET, high statistics data incorporating more counts is required. /Kir12
In this study, 3DRP method is used for image reconstruction. This process requires histogramming and histogramming requires sampling of the coincident events. Obtaining optimum sampling for better resolution is one of the main goals in the image processing. PET scanner acquires data both in axial and radial direction. Hence, optimum sampling rates for axial, radial and angular directions are required. In this study, histogramming is done with different sampling rates to see their effect in the resolution of the reconstructed images.
The construction of the scanner causes gaps and non-homogeneously sampled Filled of View (FOV). Gap filling method fill the gaps generated from inter crystal and inter module gaps. Coincident detection differs due to difference of the positions of crystals. Normalization is required to have the uniform response from all the detectors. In this study normalization and gap filling were used before 3DRP reconstruction.
Preliminary result shows the ability of resolving the capillary of 1.4 mm. But for the complex phantoms the smallest resolvable inserts was 2 mm. The best results were obtained with the sampling rate of 1mm compared to other sampling rates for different phantoms due to better statistics. The list-mode data of the phantoms used in this study is a low statistics data. To exploit the novelty of AX-PET, high statistics data incorporating more counts is required. /Kir12