Abstract
IntroductionThis study aimed first to quantify the impact of varying concentrations of iodine contrast agents on electron density (ED) values of ED maps through phantom experiments and evaluate the differences in dose calculations; Second, to validate that utilizing ED maps for photon dose calculation can minimize the uncertainties introduced by the iodine contrast agents by using actual patients' data.MethodsA phantom with iodine inserts of varying concentrations was scanned using a spectral CT to acquire post-contrast and non-contrast CT images and the corresponding ED maps. The data of 22 patients with pelvic tumors were selected retrospectively for clinical validation. Treatment plans were transplanted to non-contrast and post-contrast ED maps, and the dose distributions were then recalculated. The differences in dose-volume histogram (DVH) parameters between the plans based on non-contrast and post-contrast ED maps were compared. Besides, gamma analysis was performed to evaluate the discrepancies in dose distributions between these two plans.ResultsFor the phantom experiment, under clinical organ perfusion concentrations (≤5.0 mg/ml), the maximum increase in HU values reached 145.76, whereas the ED values showed only a 1.54% increase. Compared to the non-contrast images, the maximum dose difference of the post-contrast CT image was 2.3%, while the post-contrast ED map was only 0.2%. For comparison of treatment plans based on patients' non-contrast and post-contrast ED maps, only the bladder's D50% showed a statistically significant difference, but the difference in value remained clinically negligible. The remaining showed no statistically significant differences.ConclusionThe influence of iodine contrast agents on the ED values is minimal. Dose calculations based on ED maps can significantly reduce the errors introduced by contrast agents. This approach would allow us to depend solely on a single post-contrast scan for radiotherapy simulation, thereby minimizing imaging radiation exposure and improving simulation efficiency.