Improved fast kV-switching dual-energy CBCT with a spectral modulator: System design and a feasibility study

Fast kV-switching (FKS) has successfully been employed in commercial computed tomography (CT) systems. However, FKS in cone-beam CT (CBCT) spectral imaging still faces challenges such as scatter contamination and relatively moderate energy separation, which results in suboptimal quantitative performance. There have been some studies on using flat-field filters to improve the performance of FKS spectral imaging. However, to the best of our knowledge, few works focus on implementing a stationary but spatially varying filter to enhance the performance of FKS CBCT spectral imaging.

To improve the performance of FKS CBCT spectral imaging, we first explored the feasibility of adopting a spectral modulator (SM) in FKS CBCT spectral imaging.

In this work, we utilized the Cramér-Rao lower bound (CRLB) method to establish a parameter optimization method for FKS dual-energy CBCT with a spectral modulator, and quantitatively analyzed the impact of different spectral modulators, including materials, thickness, and spatial frequencies, on FKS spectral imaging performance. Both numerical simulations and physical experiments were conducted to validate the improvement of spectral modulator for FKS CBCT spectral imaging. In cone-beam experiments, spectral modulator with flying focal spot technology (SMFFS) was utilized to estimate scatter distribution.

In parameter optimization analysis, we found that each metallic material had an optimal thickness for each basis material image under 80/140 kVp pair scan, and the performance of FKS spectral imaging improves as spectral modulator becomes denser but with greater challenges to power capacity of x-ray tube. The copper (Cu) with thickness of 0.2 mm and molybdenum (Mo) with thickness of 0.07 mm were selected. In physical experiments, multi-energy phantom and chest phantom with a 5 mg/cc iodine rod inserted were scanned with 80/120 kVp pairs. The CNRs of water density image for multi-energy phantom in fan-beam scan were boosted by average improvements of 20.15%  $\pm$ 2.15% and 22.72%  $\pm$  0.83% for Mo modulator and Cu modulator, respectively. The CNRs of iodine density image for multi-energy phantom in fan-beam scan were boosted by average improvements of 25.31%  $\pm$ 16.93% and 30.81% $\pm$ 22.01% for Mo modulator and Cu modulator, respectively.

A new way to improve the energy separation of FKS using spectral modulator was proposed. A parameter optimization method for FKS CBCT spectral imaging with a spectral modulator was established. The feasibility of spectral modulator to improve FKS CBCT spectral imaging performance was validated.