Comparison of iodine quantification accuracy on prototype deep silicon photon-counting and energy-integrating detector CT

PURPOSE: This work examined the iodine quantification accuracy on a prototype deep silicon (dSi) photon-counting detector (PCD) computed tomography (CT) system compared to a rapid kV switching energy-integrating detector (EID) dual-energy (DE) CT system. METHODS: Iodine-containing rods (0-20 mg I/mL) in a phantom (Gammex MECT) were scanned with the prototype dSi PCD and DECT systems. Iodine (water) material density images were made with prototype and commercially available material decomposition algorithms, respectively. Circular regions of interest were placed over the center of the iodine rods to measure iodine accuracy in each slice, and slices were averaged. A correction based on the known issue of background material difference from true water was determined from the 0 mg I/mL (solid water) rod and the relative rod densities and was applied to the iodine quantification. Iodine percent error was defined as the difference between corrected iodine quantification and known iodine quantification, divided by the known iodine, multiplied by 100. RESULTS: The corrected iodine quantification was within 0.15 mg I/mL for 0-20 mg I/mL iodine rods on the prototype PCD CT and 0.2 mg I/mL for the EID DECT. This translates to iodine percent errors of 0.4-4.3% on the PCD CT and 1.1-76% on the EID DECT for 0.2-20 mg I/mL rods. CONCLUSION: Iodine quantification on the prototype dSi PCD CT was within 4.3% for all tested iodine-containing rods, which is similar or better than the performance of the EID DECT system and previous work on the prototype dSi system.