Hybrid modality dual-energy imaging aggregating complementary advantages of kV-CT and MV-CBCT: concept proposal and clinical validation

Objective. Megavoltage cone-beam CT (MV-CBCT) is advantageous in metal artifact reduction during image-guided radiotherapy (IGRT), although it is limited by poor soft tissue contrast. This study proposed and evaluated a novel hybrid modality dual-energy (DE) imaging method combining the complementary advantages of kilovoltage CT kV-CT and MV-CBCT. Approach. The kV-CT and MV-CBCT images were acquired on a planning CT scanner and a Halcyon linear accelerator respectively. After rigid registration, images of basis materials were generated using the iterative decomposition method in the volumetric images. The decomposition accuracy was quantitatively evaluated on a Gammex 1472 phantom. The performance of contrast enhancement and metal artifact reduction in virtual monochromatic images were evaluated on both phantom and patient studies. Main results. Using the proposed method, the mean percentage errors for relative electron density (RED) and stopping power ratio were 0.90% and 0.81%, outperforming the clinical single-energy mapping method with mean errors of 1.28% and 1.07%, respectively. The contrasts of soft-tissue insets were enhanced by a factor of 2-3 at 40 keV compared to kV-CT. The standard deviation in the metal artifact area was reduced by ∼67%, from 42 HU (kV-CT) to 14 HU (150 keV monochromatic). The head and neck patient test showed that the percent error of soft-tissue RED in the metal artifact area was reduced from 18.1% (HU-RED conversion) to less than 1.0% (the proposed method), which was equivalent to the maximum dosimetric difference of 28.7% based on the patient-specific plan. Significance. Without hardware modification or extra imaging dose, the proposed hybrid modality method enabled kV-MV DE imaging, providing improved accuracy of quantitative analysis, soft-tissue contrast and metal artifact suppression for more accurate IGRT.