WE‐FG‐207B‐12: Quantitative Evaluation of a Spectral CT Scanner in a Phantom Study: Results of Spectral Reconstructions

Abstract

Purpose:To evaluate the quantitative image quality of spectral reconstructions of phantom data from a spectral CT scanner.Methods:The spectral CT scanner (IQon Spectral CT, Philips Healthcare) is equipped with a dual‐layer detector and generates conventional 80ߝ140 kVp images and variety of spectral reconstructions, e.g., virtual monochromatic (VM) images, virtual non‐contrast (VNC) images, iodine maps, and effective atomic number (Z) images. A cylindrical solid water phantom (Gammex 472, 33 cm diameter and 5 cm thick) with iodine (2.0ߝ20.0 mg I/ml) and calcium (50ߝ600 mg/ml) rod inserts was scanned at 120 kVp and 27 mGy CTDIvol. Spectral reconstructions were evaluated by comparing image measurements with theoretical values calculated from nominal rod compositions provided by the phantom manufacturer. The theoretical VNC was calculated using water and iodine basis material decomposition, and the theoretical Z was calculated using two common methods, the chemical formula method (Z1) and the dual‐energy ratio method (Z2).Results:Beam‐hardening‐like artifacts between high‐attenuation calcium rods (≥300 mg/ml, >800 HU) influenced quantitative measurements, so the quantitative analysis was only performed on iodine rods using the images from the scan with all the calcium rods removed. The CT numbers of the iodine rods in the VM images (50∼150 keV) were close to theoretical values with average difference of 2.4±6.9 HU. Compared with theoretical values, the average difference for iodine concentration, VNC CT number and effective Z of iodine rods were −0.10±0.38 mg/ml, −0.1±8.2 HU, 0.25±0.06 (Z1) and −0.23±0.07 (Z2).Conclusion:The results indicate that the spectral CT scanner generates quantitatively accurate spectral reconstructions at clinically relevant iodine concentrations. Beam‐hardening‐like artifacts still exist when high‐attenuation objects are present and their impact on patient images needs further investigation.YY is an employee of Philips Healthcare.