Energy Selection to Reduce Size-Dependent HU Bias In Photon Counting CT Monoenergetic Imaging of Iodine and Calcification
Abstract
Purpose
To quantify the energy-dependent effect of phantom size on monoenergetic HU for high-density materials relevant to contrast and calcification assessment in PCCT.
Methods
A high-density insert module in a multi-energy phantom was scanned on a Siemens NAEOTOM Alpha PCCT in 20 cm and 40 cm size configurations. VMIs from 40–190 keV were reconstructed using iterative reconstruction strength. Mean HU were measured with consistent ROIs for iodine inserts with different concentrations, gadolinium, calcium/calcification surrogates (e.g., CaCO₃, cortical/inner bone), water, and background. Size dependence was quantified as ΔHU_size = HU_40cm − HU_20cm versus energy.
Results
Phantom-size dependence was strongly energy dependent and, for several materials, changed sign across the VMI range. At 40 keV, ΔHU_size was positive and largest for gadolinium (+43 HU) and cortical bone (+37 HU), with smaller but measurable positive bias for iodine (+20 HU) and water (+13 HU). At 190 keV, ΔHU_size decreased and became negative for gadolinium (−16 HU), iodine (−9 HU), and water (−9 HU); background bias was small (−3 HU). ΔHU_size decreased with keV and approached/crossed 0 in the mid-keV range for most materials, indicating ~60–80 keV minimizes size-related HU bias in this phantom (cortical bone remains positive). Across 40–190 keV, the maximum |ΔHU_size| reached ~43 HU (gadolinium), ~37 HU (cortical bone), and ~20 HU (iodine), demonstrating residual size sensitivity in quantitative monoenergetic HU despite PCCT acquisition.
Conclusion
Monoenergetic HU for contrast and calcification surrogates in PCCT remains size dependent in an energy-specific manner, with the largest bias at low keV and reduced bias at high keV. Selecting mid-keV VMIs may reduce patient/phantom size–dependent HU bias for quantitative applications. These findings support selecting energy ranges and QC expectations that account for object size when interpreting quantitative HU metrics for contrast or calcification-related tasks.