Quantification accuracy for I-123 SPECT/CT studies using LEHR and me collimators: a Monte Carlo study
| dc.contributor.advisor | Van Staden, J. A. | en_ZA |
| dc.contributor.advisor | Du Raan, H. | en_ZA |
| dc.contributor.author | Richards, Anneray | en_ZA |
| dc.date.accessioned | 2022-07-28T10:06:20Z | |
| dc.date.available | 2022-07-28T10:06:20Z | |
| dc.date.issued | 2021 | en_ZA |
| dc.description | Dissertation (MMed.Sc.(Medical Physics))--University of the Free State, 2021 | en_ZA |
| dc.description.abstract | The accurate quantification of Nuclear Medicine single photon emission tomography (SPECT) plays an important part in radiopharmaceutical therapy. Accurately quantifying SPECT images of a diagnostic radionuclide such as I-123 is desirable, although not a straightforward process as it is hindered by the complex decay scheme. I-123 has low-energy primary emissions of 159 keV, and performing acquisitions with a low-energy resolution (LEHR) collimator result in images with high resolution. However, I-123 also has high-energy photon emissions which degrade image contrast and quantification accuracy. This degradation can be reduced by using medium-energy collimators (ME); however, at the expense of spatial resolution. Most clinical facilities have access to LEHR collimators, but not necessarily ME collimators. The aim of this study was to evaluate the quantification accuracy of I-123 LEHR and ME collimated SPECT images when an optimised OSEM reconstruction protocol is applied. To accomplish the aim three objectives were identified: 1) validation of a SIMIND modelled gamma camera fitted with LEHR and ME collimators, 2) optimisation of the iterative reconstruction algorithm in terms of equivalent iterations and SPECT corrections, and based on these results, 3) evaluation of the quantification accuracy of I-123 LEHR and ME SPECT images. The first objective of this study, to validate the SIMIND modelled gamma camera fitted with LEHR and ME collimators for I-123, involved comparing measured and simulated I-123 data. Results of measured and simulated planar performance tests (system energy resolution, system spatial resolution, and system sensitivity) were compared for both collimators. The validation included a visual comparison of reconstructed SPECT images of a quality control phantom in terms of uniformity, cold contrast, resolution, and linearity. The measured and simulated planar results for system energy resolution, system spatial resolution and system sensitivity differed by 3.4%, 6.4% and 5.3%, respectively. The visual comparison performed on the reconstructed SPECT images showed good agreement between the measured and simulated data. The second objective was to optimise the OSEM iterative reconstruction algorithm concerning the number of iterations and SPECT corrections. SPECT images of voxel-based phantoms of spherical objects and image quality phantoms were simulated and reconstructed with different numbers of effective iterations. The count density recovery, image noise, contrast and resolution were evaluated. The image quality phantom was also reconstructed with different corrections (attenuation, scatter and collimator-detector response (CDR)) and compared. The optimal number of equivalent iterations was selected as 64 and the contribution of the different corrections was appreciated. When septal penetration and scatter was compensated for as part of the CDR correction, the LEHR collimator results were comparable to that obtained with the ME collimator. This led to the aim of the final objective: to determine the quantification accuracy of I-123 SPECT studies in patient phantoms acquired with LEHR and ME collimators. Using voxel based patient phantoms, the quantification accuracy was assessed for LEHR and ME SPECT images of spherical objects. Quantification errors smaller than 3.8% were recorded for both the LEHR and ME collimators when attenuation, scatter and CDR (including septal penetration and scatter) corrections were applied. Therefore, to conclude, when appropriate SPECT corrections were applied during the reconstruction of I 123 LEHR and ME SPECT images, the image quality between the collimators were comparable and quantification accuracy of up to 3.8% was achievable. | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11660/11792 | |
| dc.language.iso | en | en_ZA |
| dc.publisher | University of the Free State | en_ZA |
| dc.rights.holder | University of the Free State | en_ZA |
| dc.subject | Nuclear Medicine | en_ZA |
| dc.subject | I-123 | en_ZA |
| dc.subject | SPECT/CT | en_ZA |
| dc.subject | Monte Carlo simulations | en_ZA |
| dc.subject | LEHR collimator | en_ZA |
| dc.subject | ME collimator | en_ZA |
| dc.subject | Optimisation | en_ZA |
| dc.subject | Quantification | en_ZA |
| dc.title | Quantification accuracy for I-123 SPECT/CT studies using LEHR and me collimators: a Monte Carlo study | en_ZA |
| dc.type | Dissertation |