Comparison of three commercial multi‐energy CT and electron density phantoms for iodine quantification and tissue substitute characterization
Artur Omar, Robert Vorbau, Jens Zimmerman, Gavin PoludniowskiAbstract
Background
Multi‐energy computed tomography (MECT) and electron density (ED) phantoms are designed for evaluating the performance of CT scanners. They include substitute materials representing the energy‐ and material‐dependent CT numbers of different organs, tissues, or contrast agents. While MECT phantoms are intended for dual‐energy CT (DECT) and photon‐counting CT (PCCT) applications, ED phantoms are meant for the calibration of CT numbers to a metric relevant for radiation therapy treatment planning.
Purpose
To compare three standard commercial phantoms in terms of iodine quantification for diagnostic imaging and tissue substitute characterization for radiation therapy. The phantoms investigated include the Multi‐Energy CT/Advanced Electron Density Phantom (AED) by Sun Nuclear Corporation (Melbourne, FL, USA), the CIRS Model 662/062M phantom (Norfolk, VA, USA), and the newly introduced Multi‐Energy QA Phantom by QRM (Möhrendorf, Germany), for which we report the first evaluation results.
Methods
The phantoms were imaged using a GE Revolution Apex DECT scanner (GE Healthcare, Waukesha, WI, USA), a Siemens SOMATOM X.ceed DECT scanner (Siemens Healthineers, Forchheim, Germany), and a Siemens NAEOTOM Alpha PCCT scanner. Iodine‐in‐water inserts were evaluated using virtual monochromatic images and iodine maps, whereas the tissue equivalence of four insert sets (AED, CIRS‐ED, CIRS‐EPTN, and QRM) was theoretically assessed. For this purpose, vendor‐specified elemental compositions and physical mass densities were used to derive relative electron density (RED), effective atomic number (EAN), mean excitation energy (I‐value), and mass density (MD), which were then compared with corresponding human tissues listed in ICRU Report 44. CT scans of the phantoms containing these tissue substitutes were subsequently used for CT number calibration to proton stopping power‐ratio (SPR), RED, and MD.
Results
Fitted slopes for reconstructed iodine concentrations were accurate to within 8% for the QRM and AED phantom inserts. For the best performing scanner (PCCT), the slope was within 3%. The CIRS iodine‐in‐water inserts showed substantially higher deviations of up to 22%. Linear attenuation coefficients derived from 70 keV virtual monochromatic images generally agreed with theoretically expected values to within about 10% for the QRM and AED iodine inserts (for concentrations of 1 mg/mL or higher). Regarding the QRM, AED and CIRS‐EPTN tissue substitute sets: they were closely tissue‐equivalent in terms of RED, EAN, I‐value (and hence SPR), but not in terms of MD. The CIRS‐ED inserts were not, however, closely tissue‐equivalent in terms of EAN. Consequently, the QRM, AED and CIRS‐EPTN substitutes satisfy the consensus guide criteria of the European proton therapy community for CT calibration, whereas the CIRS‐ED set does not.
Conclusion
The commercially available phantoms of Sun Nuclear (AED) and QRM are suitable for both MECT performance assessment and CT‐calibration for radiation therapy. However, the batch of CIRS iodine‐in‐water inserts used in this study was unsuitable for MECT assessment. Also, the tissue substitute set provided with the CIRS 062M ED phantom was less tissue‐equivalent than the other sets evaluated.