How to experimentally test the accuracy of quantitative SPECT images?
Hanna Piwowarska-Bilska
Orcid-Profil, Sara Kurkowska
Orcid-Profil, Jacek Iwanowski
Orcid-Profil, Małgorzata Poniatowska
Orcid-Profil, Mateusz Piwowarski
Orcid-Profil und Bożena Birkenfeld
Orcid-Profil 
Online veröffentlicht: 11. Juni 2025
Seitenbereich: 104 - 109
Eingereicht: 05. Mai 2024
Akzeptiert: 14. Feb. 2025
DOI: https://doi.org/10.2478/pjmpe-2025-0011
Schlüsselwörter
© 2025 Hanna Piwowarska-Bilska et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Introduction
Radiopharmaceutical therapy (RPT) uses radiopharmaceuticals that target cancer cells and is increasingly used as an oncological treatment. Estimating the absorbed dose of ionizing radiation in the patient’s organs is the basis for personalized RPT. This article presents a technique for assessing the performance of a quantitative reconstruction using NEMA phantom. The goal of this work was to compare the accuracy of quantitative SPECT images using the same reconstruction algorithm but acquired with two different SPECT/CT systems from two different manufacturers.
Materials and Methods
We performed a series of SPECT/CT acquisitions of a NEMA IEC Body Phantom filled with 177Lu using two gamma cameras. The images were reconstructed and quantified using Hermes software.
Results
There was no statistically significant difference in the results of recovered activities between the Symbia Intevo Bold and the NM/CT 850 gamma cameras. For a sphere with a diameter of 37 mm, the Rc values ranged from 0.69 to 0.75 for the Symbia Intevo Bold and 0.70 to 0.73 for the NM/CT 850. For a sphere with a diameter of 28 mm, the Rc values ranged from 0.61 to 0.71 for the Symbia Intevo Bold and 0.61 to 0.70 for the NM/CT 850. For a sphere with a diameter of 22 mm, the Rc values fell within the range of 0.56 to 0.63 for the Symbia Intevo Bold and 0.51 to 0.59 for the NM/CT 850.
Conclusions
Quantitative SPECT images are necessary to perform dosimetry calculations and before using them in the clinical settings it is important to perform validation. This validation can be simply performed by acquiring images of the phantom filled with known activity using the same acquisition and reconstruction parameters as those employed for patients studies. In our study we observed comparable performance between images acquired by the systems from two different manufacturers.