Evaluation 99mTc and 123I quantification using SPECT/CT

[摘要] English: A review of Single Photon Emission Tomography (SPECT) quantification shows thatdifferent protocols and phantoms are available to evaluate SPECT quantificationaccuracy. This study was necessitated by the lack of standardized protocols and thewidespread use of a variety of non-standard phantoms. The aim of this work was toevaluate the influence of the geometry of a radionuclide distribution on SPECTquantification accuracy for 99mTc and 123I isotopes in an abdominal phantom. In order toachieve the aim, the following steps were taken: The preparatory phase of the studywas to design and construct an abdominal phantom, verify the accuracy of theattenuation coefficients obtained with the Computed Tomography (CT) scanner,determine the accuracy of the source calibrator used in this study and then obtain acalibration factor in order to convert image counts to activity. During the quantificationphase SPECT data were acquired, the influence of not applying scatter correctionexplicitly was evaluated and the final quantification was performed using the proposedstandard clinical reconstruction protocols. The influence of different tumour sizes andlocations in the abdominal phantom relative to a high uptake organ on the quantificationaccuracy was evaluated. Finally, parameters in the Ordered Subset ExpectationMaximization (OSEM) reconstruction protocol were altered in order to investigate theinfluence of number of subsets and iterations on the quantified data.The non-standard Density Phantom with five different compounds was used for theverification of the 99mTc and 123I attenuation coefficients. The percentage differencebetween the measured and theoretical attenuation coefficients values were < 3%,except for Polystyrene (85% and 65% respectively). The SPECT calibration factor wasdetermined for both 99mTc (11.0 ± 1.3 cpm/kBq) and 123I (10.8 ± 0.3 cpm/kBq) using theCylindrical Phantom. The in-house built Abdominal Phantom was used to evaluate thetumour activity quantification accuracy. The quantification accuracy of 99mTc and 123Iwas found to change significantly (p < 0.05) as a function of tumour size aftercorrections for 'spill out counts due to the partial volume effect, scatter and attenuation were applied. On the other hand, there was no significant difference in the quantificationaccuracy (p > 0.05) for each tumour at different tumour-liver distances when appropriatescatter and attenuation corrections were applied. The influence of OSEM parametersshowed no dependence on the tumour-liver distance and no significant difference(p > 0.05) between quantification with background activity as compared to nobackground activity.In conclusion, the study showed that the quantification accuracy for 99mTc and 123I wascomparable to other published studies. It was found that the tumour quantificationaccuracy is not influenced by proximity of high uptake organs when appropriatecorrection factors were applied. Tumour size influenced the accuracy of SPECTquantification for both radionuclides. The results of this study also showed that at least128 Maximum Likelihood Expectation Maximization (MLEM) equivalent iterations wereneeded during iterative reconstruction to achieve convergence and consistent SPECTquantification accuracy. Finally, it is recommended that the evaluated quantificationprotocol may be used in our nuclear medicine clinic for 99mTc and 123I quantification.

[发布机构] University of the Free State