Assessment of collimators in radium-223 imaging with channelized Hotelling observer: a simulation study
Radium-223 (223Ra) is used in unsealed radionuclide therapy for metastatic bone tumors. The aim of this study is to apply a computational model observer to 223Ra planar images, and to assess the performance of collimators in 223Ra imaging.
The 223Ra planar images were created via an in-house Monte Carlo simulation code using HEXAGON and NAI modules. The phantom was a National Electrical Manufacturers Association body phantom with a hot sphere. The concentration of the background was 55 Bq/mL, and the sphere was approximately 1.5–20 times that of the background concentration. The acquisition time was 10 min. The photopeaks (and the energy window) were 84 (full width of energy window: 20%), 154 (15%), and 270 keV (10%). Each 40 images, with and without hot concentration, were applied to a three-channel difference-of-Gaussian channelized Hotelling observer (CHO), and the signal-to-noise ratio (SNR) of the hot region was calculated. The images were examined using five different collimators: two low-energy general-purpose (LEGP), two medium-energy general-purpose (MEGP), and one high-energy general-purpose (HEGP) collimators.
The SNR value was linearly proportional to the contrast of the hot region for all collimators and energy windows. The images of the 84-keV energy window with the MEGP collimator that have thicker septa and larger holes produced the highest SNR value. The SNR values of two LEGP collimators were approximately half of the MEGP collimators. The HEGP collimator was halfway between the MEGP and LEGP. Similar characteristics were observed for other energy windows (154, 270 keV). The SNR value of images captured via the 270-keV energy window was larger than 154-keV, although the sensitivity of the 270-keV energy window is lower than 154-keV. The results suggested a positive correlation between the SNR value and the fraction of unscattered photons.
The SNR value of CHO reflected the performance of collimators and was available to assess and quantitatively evaluate the collimator performance in 223Ra imaging. The SNR value depends on the magnitudes of unscattered photon count and the fraction of unscattered photon count. Consequently, in this study, MEGP collimators performed better than LEGP and HEGP collimators for 223Ra imaging.
KeywordsRadium-223 Monte Carlo simulation Channelized Hotelling observer Radionuclide therapy Bone scintigraphy
The authors would like to thank Dr. Shuzo Uehara for providing the original simulation codes (HEXAGON and NAI) and a wealth of valuable advice. The authors would also like to thank Mr. Ryota Oshima for his assistance. We have no conflicts of interest to disclose.
- 3.Carrasquillo JA, O’Donoghue JA, Taskar NP, Humm JL, Rathkopf DE, Slovin SF, et al. Phase I pharmacokinetic and biodistribution study with escalating doses of 223Ra-dichloride in men with castration-resistant metastatic prostate cancer. Eur J Nucl Med Mol Imaging. 2013;40:1384–93.CrossRefGoogle Scholar
- 5.Chittenden SJ, Hindorf C, Parker CC, Lewington VJ, Pratt BE, Johnson B, et al. A phase 1, open-label study of the biodistribution, pharmacokinetics, and dosimetry of 223Ra-dichloride in patients with hormone-refractory prostate cancer and skeletal metastases. J Nucl Med. 2015;56:1304–9.CrossRefGoogle Scholar
- 6.Pacilio M, Ventroni G, Cassano B, Ialongo P, Lorenzon L, Castro ED, Recine F, Sternberg CN, Mango L. A case report of image-based dosimetry of bone metastases with Alpharadin (223Ra-dichloride) therapy: inter-fraction variability of absorbed dose and follow-up. Ann Nucl Med. 2016;30:163–8.CrossRefGoogle Scholar
- 23.Li K, Garrett J, Ge Y, Chen G-H. Statistical model based iterative reconstruction (MBIR) in clinical CT systems. Part II. Experimental assessment of spatial resolution performance. Med Phys. 2014;41:071911-1–12.Google Scholar