Motion blurring artifacts in CBCT can be alleviated by providing a sequence of phase-depended images through 4DCBCT technique. However, it introduces streaking artifacts due to the under-sampled projection problem for each phase. One possible solution is to use deformable registration algorithms to estimate the deformation vector fields (DVF) between different phase-depended images. Among them, the optical flow based Demons registration method is a major technique due to its simplicity and efficiency. However, current Demons algorithms still suffer from relative low registration precision due to only using gradient information of images to calculate the DVFs in different directions. To improve the registration precision, we took the interaction between the DVFs calculated in Demons process into account and then proposed a weighted Demons registration method. In this method, a joint distribution of the gradient magnitude and Laplace of Gaussian (GM-LoG) signal which could represent the edge features of magnitude and orientation was introduced. Such a joint distribution could be used to guide the calculation of DVF to preserve the more detailed features and topology structure of the image during the registration process. Both simulation and real data experiments have been carried out to verify the performance of our method. In specific, the image quality has been improved regarding to distinct features, especially in regions of interest of moving tissues. Quantitative evaluations were shown in terms of the rooted mean square error (RMSE) and correlation coefficient (CC) are achieved by our method when compared with existing single Demons method and double Demons method, respectively.
Regularization parameter selection is pivotal in optimizing reconstructed images which controls a balance between fidelity and penalty term. Images reconstructed with the optimal regularization parameter will keep the detail preserved and the noise restrained at the same time. In previous work, we have used CT image statistics to select the optimal regularization parameter by calculating the second order derivates of image variance (Soda-curve). But same as L-curve method, it also needs multiple reconstruction in different regularization parameters which will spend plenty of time. In this paper, we dive into the relationship between image statistics changes and regularization parameter during the iteration. Meanwhile, we propose a method based on the empirical regularity found in the iterations to tune the regularization parameter automatically in order to maintain the image quality. Experiments show that the images reconstructed with the regularization parameters tuned by the proposed method have higher image quality as well as less time when compared to L-curve based results.
Conventional Cone-Beam Computed Tomography (CBCT) acquisition suffers from motion blurring artifacts at the region of the thorax, and consequently, it may result in inaccuracy in localizing the target of treatment and verifying delivered dose in radiation therapy. Although 4D-CBCT reconstruction technology is available to alleviate the motion blurring artifacts with the strategy of projection sorting followed by independent reconstruction, under-sampling streaking artifacts and noise are observed in the set of 4D-CBCT images due to relatively fewer projections and large angular spacing in each phase. Aiming at improving the overall quality of 4D-CBCT images, we explored the performance of the deep learning model on 4D-CBCT images, which has been paid little attention before. Inspired by the high correlation among the 4D-CBCT images at different phases, we incorporated a prior image reconstructed from full-sampled projections beforehand into a lightweight structured convolutional neural network (CNN) as one input channel. The prior image used in the CNN model can guide the final output image to restore detailed features in the testing process, so it is referred to as Prior-guided CNN. Both simulation and real data experiments have been carried out to verify the effectiveness of our CNN model. Experimental results demonstrate the effectiveness of the proposed CNN regarding artifact suppression and preservation of anatomical structures. Quantitative evaluations also indicate that 33.3% and 21.2% increases in terms of Structural Similarity Index (SSIM) have been achieved by our model when comparing with gated reconstruction and images tested on CNN without prior knowledge, respectively.
Conventional Cone-Beam Computed Tomography (CBCT) acquisition suffers from motion blurring problems of moving organs, especially the respiratory motion at thorax region, and consequently it may result in inaccuracy in the localizing the target of treatment and verifying delivered dose in radiation therapy. Although 4D-CBCT reconstruction technology is available to alleviate the motion blurring artifacts with the strategy of projection sorting tuned by respiratory bins, it introduces under-sampled problems. Aiming to precisely estimate the motion information of individual 4D-CBCT reconstructions, the proposed method combines the motion variable matrixes extracted from independent 4D-CBCT reconstructions using Robust Principal Component Analysis (RPCA) and the prior reconstructed image from fullsampled projections together and incorporate into iterative reconstruction framework, defining the Motion Compensated RPCA (MC-RPCA) method. Both simulation data and real data have been tested to verify the improvement in image quality at individual reconstructed phases by MC-RPCA. It can be obviously observed that the image quality the MC-RPCA method is improved with distinct features, especially in two regions of interest (ROI) with moving tissues. Quantitative evaluations indicate that large improvements in the Structural Similarity Index (SSIM) and Contrast-to-Noise Ratio(CNR) are achieved at the diaphragm slice by our method when comparing with MKB and the Prior Image Constraint Compressed Sensing (PICCS) algorithm, respectively.
Digital breast tomosynthesis (DBT) can provide quasi three-dimensional (3D) structural information using a sequence of projection views that are acquired at a small number of views over a limited angular range. Nevertheless, the quantitative accuracy of the image can be significantly compromised by severe artifacts and poor resolution in depth dimension resulting from the incomplete data. The purpose of this work is: (a) investigate a variety of boundary artifacts representing as the decline tendency of the attenuation coeﬃcients which is caused by insuﬃcient projection data; (b) employ the 3D breast surface information we proposed in this study into the simultaneous algebraic reconstruction technique (SART) for artifacts reduction. Numerical experiments demonstrated that such boundary artifacts could be suppressed with the proposed algorithm. Compared to SART without using prior information, a 9.57% decrease in root mean square error (RMSE) is achieved for the central 40 slices. Meanwhile, the spatial resolution of potential masses and micro calcifications (MCs) in the reconstructed image is relatively enhanced. The full-width at half maximum (FWHM) of the artifact spread function (ASF) for proposed algorithm and SART are 17.87 and 19.68, respectively.