Dual-energy cone-beam CT improves iodine contrast and Hounsfield unit value stability in the maxillofacial region: A phantom study

This study was designed to assess the feasibility of dual-energy cone-beam CT (DE-CBCT) and the associated image processing algorithms to improve delayed phase iodine enhancement. An anthropomorphic adult head phantom with iodine inserts in maxillofacial regions was imaged using a clinical CBCT and a helical CT scanner. DE-CBCT was acquired sequentially at 80 and 120 kV, and helical CT at 100 kV (routine clinical protocol), with radiation doses matched within 5%. Virtual monochromatic images (VMIs) at 50 and 200 keV were generated with (VMI+) and without (VMI) noise-optimized frequency splitting algorithm. Evaluation included Hounsfield Unit (HU) agreement with helical CT, iodine contrast, contrast-to-noise ratio (CNR), artifact reduction, and subjective radiologist assessment of target visibility. VMI50 and VMI50+ demonstrated the strongest HU agreement with CT. VMI50+ produced significantly better contrast than CBCT in the maxillary sinus region (mean percentage change 40.7% (15.3-66.1%); p < 0.05), while improvements were minor in the mandibular region (9.7% (3.0-16.5%); p < 0.05). VMI50+ significantly increased CNR in the maxillary sinus compared with CBCT (3.3 vs. 2.8; p < 0.05), with no significant difference in the mandibular region. Although DE-CBCT increased contrast and reduced artefacts, it did not reveal any additional targets compared with routine CBCT. DE-CBCT with VMI and frequency splitting improves HU agreement with CT, slightly improves soft tissue visualization, and reduces artifacts in head and neck phantom imaging.