Room: Exhibit Hall | Forum 9
Purpose: To automate the window level setting (W/L) adjustment for iodine enhanced virtual monoenergetic images (VMIs) with varying mono-energy in dual energy CT.
Methods: We first derived a simplified formula for Hounsfield Unit as consisting of two separate components: an energy specific factor Q(E) and a material specific parameter pair (Ï?', Z'). To determine the Q(E) factor, the Gammex-467 CT phantom was used as the calibration phantom with known material composition and measured HU(E) values on VMIs. As iodine contrast is usually used for enhancing soft tissues such as brain, liver and etc., the material parameter space (Ï?', Z') was modeled using five points: water and iodinated contrast with four different iodine concentration (ISOVUE-200, ISOVUE-250, ISOVUE-300 and ISOVUE-370). The W/L settings for VMIs at different energy were mapped from the reference conventional W/L settings established for 120kVp, based on the HU value conversion on the material parameter space.
Results: A linear relationship was found to well fit the five points in the material specific parameter pair (Ï?', Z') space such that a singular parameter could be used as representative of the material information when iodine contrast was applied. For conventional liver contrast W/L = 150/30, the clinically preferred VMI at 40keV with highest intrinsic image contrast has the corresponding W/L = 596/111. The W/L adjustments for other mono-energies were also obtained. For conventional brain contrast W/L = 80/40, the corresponding W/L for VMI at 40keV was found as 318/151.
Conclusion: The increased HU variation with iodine enhancement at lower energies is properly accounted for with the proposed method for W/L adjustments on VMIs. Automation in W/L adjustments for VMIs helps reduce the effort of image readers for optimal image perception.