Purpose: Dual-energy CT (DECT) strengthens the material characterization and quantification due to its ability of material decomposition. The image-domain multi-material decomposition (MMD) via matrix inversion suffers from serious degradation of the signal-to-noise ratios of the decomposed images, and thus the clinical application of DECT is limited. Based on the analysis of the noise propagation of MMD, the noise in the decomposed images conforms to the highly correlated and elliptical Gaussian distribution and only distributes in two perpendicular directions. We propose an effective MMD noise suppression algorithm to improve the material-decomposed image quality by finding the principal axis along which the noise perturbation is minimal.
Methods: The noise is suppressed along the principal axis by estimating the center-of-mass value of the same-material pixel group. The proposed method is evaluated on the line-pair and contrast-rod slices of the CatphanÂ©600 phantom and one pelvic patient data. We apply the direct inversion and the block-matching and three-dimensional (BM3D) filtration methods for results comparison.
Results: The results of CatphanÂ©600 phantom and the pelvic patient show that the proposed method successfully suppresses the noise of the basis material images by one order of magnitude and preserve the spatial resolution and the noise power spectrum of the decomposed images. Compared with the BM3D filtration method, the proposed method maintains the texture distribution of the decomposed images at the same signal-to-noise ratio and the accuracy of the electron density measurement.
Conclusion: The achieves excellent noise suppression compared to the BM3D filtration and direct inversion methods. The spatial distribution is maintained when calculating the noise power spectrum of the material-decomposed images using the proposed method. The results indicate that the proposed method has the potential to be applied to clinical practice due to its increased decomposition accuracy and suppressed noise.