Photon-counting detectors with pulse-height analysis have shown promise for improved spectral CT imaging. This study investigated a novel spectral CT reconstruction method that directly estimates basis-material images from the measured energy-bin data (i.e., ‘one-step’ reconstruction). The proposed algorithm can incorporate constraints to stabilize the reconstruction and potentially reduce noise. The algorithm minimizes the error between the measured energy-bin data and the data estimated from the reconstructed basis images. A total variation (TV) constraint was also investigated for additional noise reduction. The proposed one-step algorithm was applied to simulated data of an anthropomorphic phantom with heterogeneous tissue composition. Reconstructed water, bone, and gadolinium basis images were compared for the proposed one-step algorithm and the conventional ‘two-step’ method of decomposition followed by reconstruction. The unconstrained algorithm provided a 30% to 60% reduction in noise standard deviation compared to the two-step algorithm. The fTV =0.8 constraint provided a small reduction in noise (∼ 1%) compared to the unconstrained reconstruction. Images reconstructed with the fTV =0.5 constraint demonstrated 77% to 94% standard deviation reduction compared to the two-step reconstruction, however with increased blurring. There were no significant differences in the mean values reconstructed by the investigated algorithms. Overall, the proposed one-step spectral CT reconstruction algorithm provided three-material-decomposition basis images with reduced noise compared to the conventional two-step approach. When using a moderate TV constraint factor (fTV = 0.8), a 30%-60% reduction in noise standard deviation was achieved while preserving the edge profile for this simulated phantom.
[1]
Antonin Chambolle,et al.
A First-Order Primal-Dual Algorithm for Convex Problems with Applications to Imaging
,
2011,
Journal of Mathematical Imaging and Vision.
[2]
J. Schlomka,et al.
Experimental feasibility of multi-energy photon-counting K-edge imaging in pre-clinical computed tomography
,
2008,
Physics in medicine and biology.
[3]
E. Sidky,et al.
Convex optimization problem prototyping for image reconstruction in computed tomography with the Chambolle–Pock algorithm
,
2011,
Physics in medicine and biology.
[4]
Katsuyuki Taguchi,et al.
A cascaded model of spectral distortions due to spectral response effects and pulse pileup effects in a photon-counting x-ray detector for CT.
,
2014,
Medical physics.