Color K-edge imaging using spectral photon-counting CT with high-Z elements for lung functional imaging: A phantom study.

PURPOSE: To assess the feasibility and performance of color K-edge imaging using spectral photon-counting CT (SPCCT) combined with eight high-Z-elements candidates for lung functional imaging.

MATERIAL AND

METHODS

Solutions of gadolinium, holmium, ytterbium, hafnium, tantalum, tungsten, gold, and bismuth, were prepared in 12 tubes from 0 to 2 mg/mL, and inserted into an anthropomorphic lung phantom (QRM GmbH). Scans were performed using lung protocols, on a clinical SPCCT (Philips, Israel), at 120 kVp and 3 radiation exposure levels (50, 75, 150 mAs). Noise and contrast-to-noise ratios (CNR) were measured in all images as well as relative percentage errors (RPE). Linear regressions (slope, intercept) of measured versus prepared concentrations and CNR versus concentration were plotted..

RESULTS

On average, mean absolute RPE ranged from 15.9%±8.0% for ytterbium to 34.7%±19.0% for bismuth. Quantification accuracy was excellent for all elements but lower for bismuth. For all elements and exposure levels, mean noise in color K-edge and conventional images was 0.030 ± 0.007 mg/mL and 20.4 ± 2.7 HU, respectively. CNR in color K-edge images increased as the concentrations increased, with an average slope (a) of 43.2 ± 11.8 mL/mg, the highest being for ytterbium (60.1 ± 8.1 mL/mg) and the lowest for bismuth (23.8 ± 5.2 mL/mg). CNR in conventional images also increased as the concentrations increased, with an average slope of 1.7 ± 0.5 mL/mg; the highest was for gadolinium (2.3 ± 0.3 mL/mg) and the lowest for tungsten (1.2 ± 0.1 mL/mg)..

CONCLUSION

Color K-edge imaging is feasible with multiple high-Z elements, whilst outperforming the performance of conventional imaging, paving the way for lung SPCCT functional imaging.