PT  - JOURNAL ARTICLE
AU  - Esther Vicente
AU  - Steven Rowe
AU  - Martin Lodge
AU  - Richard Wahl
AU  - Eric Frey
TI  - Investigation of volume-of-interest (VOI) definition methods for estimating organ activity concentrations in quantitative SPECT (QSPECT)
DP  - 2015 May 01
TA  - Journal of Nuclear Medicine
PG  - 1278--1278
VI  - 56
IP  - supplement 3
4099  - http://jnm.snmjournals.org/content/56/supplement_3/1278.short
4100  - http://jnm.snmjournals.org/content/56/supplement_3/1278.full
SO  - J Nucl Med2015 May 01; 56
AB  - 1278 Objectives Estimates of organ activity concentration are essential in dosimetry and may be useful for normalization in quantitative tumor imaging. The gold standard for determining mean organ activity concentration is using 3D whole-organ VOIs. However, defining full 3D organ VOIs can be time consuming. We thus investigated two alternatives: estimating organ activity concentrations from an anatomic VOI in a single transaxial slice and from a sphere placed inside the organ boundaries.Methods Low-noise Monte Carlo simulated projections were generated based on patient In-111 Octreoscan attenuation maps and organ activities and boundaries. We modeled constant activity concentrations in each of the liver, spleen, kidneys, and background. Images were reconstructed using previously-validated QSPECT methods. We compared activity concentration estimates obtained using the 3 VOI definition methods mentioned above. Various levels of morphological erosion of the true organ VOIs were used as constraints on VOI placement to investigate the effects of placing VOIs near organ boundaries.Results Biases in activity concentrations using the true 3D VOIs were lower than 5% for the liver and negligible when excluding organ boundary voxels. Activity was underestimated in whole-organ VOIs for kidneys due to partial volume effects. Bias was reduced to < 5% if VOIs were 1 voxel eroded from the true organ boundaries. The biases were similar for the noise levels studied. Biases of <10% were achievable for single-slice VOIs for slices away from organ extremities. Biases < 10% were achievable for spheres of radii > 9 mm and positioned away from boundaries, but sphere position was not otherwise critical.Conclusions Organ activity concentration with biases better than 10% was feasible without detailed definition of the full organ VOI in cases where the activity concentration is uniform. As long as positioned away from organ boundaries, the positions of the simple organ VOIs were not critical.Research Support This work was supported by the National Cancer Institute of the National Institutes of Health under award numbers U01 CA140204 and R01 CA 109234. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.