Iodine Quantification Using Dual-Energy Multidetector Computed Tomography Imaging Phantom Study Assessing the Impact of Iterative Reconstruction Schemes and Patient Habitus on Accuracy

Purpose: The aim of this study was to assess the accuracy of iodine quantification based on spectral dual-energy computed tomography (DECT) extraction with additional noise reduction using iterative reconstruction in simulated optimal and obese patient environments. Materials and Methods: Two custom-designed DECT phantoms were containing 10 vials with iodine concentrations representing arterial/parenchymal enhancement ranging from water isodensity to similar to 150 Hounsfield units and, in addition, 40 vials simulating enhancement seen in nondiluted thoracic inlet vasculature and urinary bladder/renal collecting systems of up to similar to 2000 Hounsfield units. Dual-energy computed tomography acquisition was performed using a dual-source scanner at 140 kVp/90 mAs and 80 kVp/495 mAs. Backprojection-based soft tissue kernels and corresponding iteratively reconstructed kernels generated dual-energy series used for iodine extraction. Fractional variations between known and spectrally determined iodine concentration were calculated for each concentration step; paired t tests evaluated variations between backprojected and iteratively reconstructed data sets for small and obese phantoms. Bland-Altman plots with regression analyses assessed concentration differences observed in backprojected and iteratively reconstructed data. Results: For backprojected data, mean concentration variations of 8.7% +/- 8.4 and 12.2% +/- 6.3 were detected in small and large phantoms, respectively, compared with significantly less variation observed in iteratively reconstructed data with 6.1% +/- 6.2 and 11.0% +/- 6.5, respectively. Dual-energy quantification systematically overestimated concentrations in lower concentration ranges and underestimated concentrations in higher concentration ranges. Regression analyses showed cubic distribution of concentration differences for backprojected (R-2 = 0.697) and linear distribution for iteratively reconstructed data (R-2 = 0.701). Conclusion: Spectral DECT-based iodine quantification is able to accurately quantify iodine in phantoms simulating optimal and large patients; iterative reconstruction improves the accuracy of iodine detection. Systematic deviations of the spectrally determined iodine concentrations could potentially be corrected with weighting curves.