The effect of tumour geometry on the quantification accuracy of 99mTc and 123I in planar phantom images

English: Accurate activity quantification is important for its application in radiation dosimetry. Planar image quantification plays an important role in the quantification of whole body images which provide a full assessment of bio-distribution from radionuclide administrations. In the Department of Nuclear Medicine at Universitas Hospital, 123I meta-iodobenzylguanidine [123I]-MIBG quantification of neuroendocrine tumours is performed prior to therapeutic radionuclide treatment. The bio-distribution of activity in these studies is mostly in the abdominal region. Factors influencing quantification include scatter, attenuation, background activity and close proximity of organs with radioactivity uptake. The aim of this study was to evaluate the effect of tumour geometry on the quantification accuracy of 99mTc and 123I in planar phantom images, by applying scatter and attenuation corrections, with the focus on neuroendocrine tumours. The tumour geometry investigated included: various tumour sizes, various tumour-liver distances and two tumour-background ratios (0.5 % and 1.0 %). The quantification technique was first developed with the readily available 99mTc and subsequently applied to the more costly 123I used for imaging neuroendocrine tumours. Adjustments were necessary due to the difference in physical properties between the two isotopes. An in-house manufactured abdominal phantom was developed to mimic the clinical geometries under investigation. The phantom was equipped with cylindrical inserts used to simulate tumours (diameters of the tumours were 63 mm, 45 mm, 34 mm and 23 mm) and a slider to vary the tumour-liver distance. The processing technique incorporated the use of the geometric mean method with corrections for scatter and attenuation performed on image counts. Scatter correction was performed using a modified triple energy window scatter correction technique for 99mTc and 123I, according to gamma camera manufacturer specifications. Attenuation correction was performed using transmission images obtained with an uncollimated 99mTc printed source. Scatter contribution from the abdominal phantom and transmission source combination was limited by setting the detector transmission source distance to 73 cm. A system calibration factor, processed in the same manner as the tumour quantified data was used to convert the image counts to units of radioactivity. Partial volume effect (PVE), was compensated for by the manner in which regions for tumour activity distribution were defined. The activity measured in the dose calibrators served as a reference for determining the accuracy of the quantification. The largest percentage deviation was obtained for the smallest tumours. The average activity underestimations were 29.2 ± 1.3 % and 34.6 ±1.2 % for 99mTc and 123I respectively. These large underestimations observed for the smallest tumours were attributed to PVE, which diminished with increasing tumour sizes. Better quantification accuracy was observed for the largest tumour with overestimations of 3.3 ± 2.6 % and 3.1 ± 3.0 % for 99mTc and 123I respectively. PVE compensation resulted in improved quantification accuracy for all tumour sizes yielding accuracies of better than 9.1 % and 12.4 % for 99mTc and 123I respectively. Scatter contribution to the tumours from the liver had minimal effect on the quantification accuracy at tumour-liver distances larger than 3 cm. An increased tumour-background ratio resulted in an increase in the quantification results of up to 16.6 % for calculations without PVE compensation. This contribution was increased to 26.3 % when PVE were compensated for, using larger regions. The literature often report accurate planar quantification results, however, this study shows that it is important to consider the specific tumour geometry for the study. It remains the responsibility of the user to evaluate the clinical available software and implement it in a responsible manner. When applying all relevant corrections for scatter, attenuation and PVE without significant background, quantification accuracy within 12 % was obtained. This study has demonstrated successful implementation of a practical technique to obtain planar quantitative information.