h1

TY  - THES
AU  - Issa, Ahlam Said Mohamad
TI  - A detector block-pairwise dead time correction method for improved quantitation accuracy for a dedicated BrainPET scanner
PB  - Rheinisch-Westfälische Technische Hochschule Aachen
VL  - Dissertation
CY  - Aachen
M1  - RWTH-2024-00241
SP  - 1 Online-Ressource : Illustrationen
PY  - 2023
N1  - Veröffentlicht auf dem Publikationsserver der RWTH Aachen University 2024
N1  - Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023
AB  - A Detector Block-Pairwise Dead Time Correction Method for Improved Quantitation Accuracy for a Dedicated BrainPET Scanner Dead time correction (DTC) is of high significance for accurate quantification in PET, just like other corrections for attenuation, decay, and scatter. Many PET systems use the global DTC, i.e., an average DTC factor is computed for all scintillation detector blocks of the system. However, the count rates of the individual scintillation detector blocks are potentially very different due to the individually varying irradiation of each block detector, especially for systems where the allocation of radiation shields is not possible, as in the case of our dedicated Siemens 3T MR BrainPET insert. For that reason, we have developed a block-pairwise DTC. In our approach, we extended a previously published method that uses the delayed random coincidence count rate to estimate the dead time in the individual blocks and planes. This DTC was evaluated with decay experiments using phantom measurements with homogenous and inhomogeneous activity concentrations and with and without out-of-FOV activity. We compared the accuracy and the noise behavior with measurements using a 3-compartment phantom. Moreover, we showed that the global and the improved block-pairwise DTC require different calibration. Therefore, we cross-calibrated both methods against each other. The differences in the quantification of the BrainPET images were evaluated by using several radioactive tracers. For this, we validated the method by quantifying the impact on [11C]ABP688 time-activity curves (TACs) and derived quantities such as the non-displaceable binding potential (BPND) and the total distribution volume (VT). We further studied the new method’s impact on O-(2-[18F]fluoroethyl)-L-tyrosine (FET) TACs and tumor to background ratios (TBRmax and TBRmean) and we evaluated the impact on [15O]H2O TACs and the rate constants K1 and k2, the regional cerebral blood flow (rCBF), and the VT obtained by kinetic modeling. The phantom measurements showed that the global DTC led to significant quantification biases in mainly those regions with high activity concentrations, while the block-pairwise DTC led to substantially less bias. The noise level was comparable for both methods. The evaluation of typical applications in volunteer and patient measurements revealed relevant differences between the two DTC, particularly relevant for research applications in neuroscientific studies. In case of PET imaging with [11C]ABP688 we found a relevant bias of VT in all studied brain regions when using the global DTC. For [18F]-FET-PET, differences in TBRmax of up to 10
LB  - PUB:(DE-HGF)11
DO  - DOI:10.18154/RWTH-2024-00241
UR  - https://publications.rwth-aachen.de/record/976443
ER  -