Premium
Quantitative 90 Y image reconstruction in PET
Author(s) -
Willowson Kathy,
Forwood Nicholas,
Jakoby Bjoern W.,
Smith Anne M.,
Bailey Dale L.
Publication year - 2012
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4762403
Subject(s) - imaging phantom , nuclear medicine , iterative reconstruction , positron emission tomography , partial volume , physics , medical imaging , materials science , mathematics , biomedical engineering , medicine , computer science , artificial intelligence
Purpose: Positron emission tomography (PET) imaging is increasingly used to confirm localization of 90 Y microspheres in the treatment of liver cancer. The aim of this work was to evaluate the quantification of 90 Y PET data on a current generation time‐of‐flight extended axial field‐of‐view PET/CT camera. Methods: The International Electrotechnical Commission (IEC) body phantom was used to image six spheres of varying diameters containing a high concentration of 90 Y solution in a lower concentration background. Multiple PET studies were acquired of the phantom over a number of days during decay. The effect of reconstruction parameters in OSEM was evaluated both qualitatively and quantitatively. Expected values of total phantom activity, hot‐sphere, and background concentration were compared to measured values from the reconstructed data as well as misplaced events in a cold insert. The partial volume effect was measured and the effects of time‐of‐flight during reconstruction on hot contrast recovery and background variability were evaluated according to NEMA‐NU2‐2007 protocol, and compared to that for 18 F. The method was applied to a patient study following radioembolization to estimate actual implanted radioactivity. Results: Increasing the number of OSEM iterations visually deteriorated image data and resulted in a larger overall difference of hot concentration measures when considering both count high and count poor data. The average difference between measured and true total activity and background concentration was found to be +5% and +5%, respectively. Measured hot‐sphere concentration was linear across all datasets, and while estimated to be within error of expected values, was consistently underestimated by an average of 23%, 12%, and 8%, when using a CT‐derived, 50% threshold‐derived, and 70% threshold‐derived volume of interest, respectively. Partial volume effects were evident in all but the largest sphere, following an expected relationship between object size and recovery coefficient, inferior to that of 18 F. Time‐of‐flight improved contrast of hot‐spheres but resulted in a deterioration of background variability, following a similar trend to that seen with 18 F. The patient data estimated a total implanted activity of 1643 MBq, compared to the intended dose of 1780 MBq, with a difference most likely due to residual and error in the initial dose calibration. Conclusions: Quantitative 90 Y PET with a state‐of‐the‐art PET/CT scanner with time‐of‐flight and standard corrections for photon interactions demonstrates consistent and acceptable measures of total activity and radionuclide concentration across a range of realistic count statistics. The method is suitable for measuring the radioactivity delivered at the time of 90 Y therapy with the potential for absorbed dose calculation.