Sparse-view CT is a promising approach in reducing the X-ray radiation dose in clinical CT imaging. However, the CT images reconstructed from the conventional filtered backprojection (FBP) algorithm suffer from severe streaking artifacts. Iterative reconstruction (IR) algorithms have been widely adopted to mitigate these streaking artifacts, but they may prolong the CT imaging time due to the intense data-specific computations. Recently, model-driven deep learning (DL) CT image reconstruction method, which unrolls the iterative optimization procedures into the deep neural network, has shown exciting prospect in improving the image quality and shortening the reconstruction time. In this work, we explore the generalized unrolling scheme for such iterative model to further enhance its performance on sparse-view CT imaging. By using it, the iteration parameters, regularizer term, data-fidelity term and even the mathematical operations are all assumed to be learned and optimized via the network training. Results from the numerical and experimental sparse-view CT imaging demonstrate that the newly proposed network with the maximum generalization provides the best reconstruction performance.
Background: Magnetic resonance (MR) quantitative T 1ρ imaging has been increasingly used to detect the early stages of osteoarthritis. The small volume and curved surface of articular cartilage necessitate imaging with high in-plane resolution and thin slices for accurate T 1ρ measurement. Compared with 2D T 1ρ mapping, 3D T 1ρ mapping is free from artifacts caused by slice cross-talk and has a thinner slice thickness and full volume coverage. However, this technique needs to acquire multiple T 1ρ -weighted images with different spin-lock times, which results in a very long scan duration. It is highly expected that the scan time can be reduced in 3D T 1ρ mapping without compromising the T 1ρ quantification accuracy and precision.Methods: To accelerate the acquisition of 3D T 1ρ mapping without compromising the T 1ρ quantification accuracy and precision, a signal-compensated robust tensor principal component analysis method was proposed in this paper. The 3D T 1ρ -weighted images compensated at different spin-lock times were decomposed as a low-rank high-order tensor plus a sparse component. Poisson-disk random undersampling patterns were applied to k-space data in the phase-and partition-encoding directions in both retrospective and prospective experiments. Five volunteers were involved in this study. The fully sampled k-space data acquired from 3 volunteers were retrospectively undersampled at R=5.2, 7.7, and 9.7, respectively.Reference values were obtained from the fully sampled data. Prospectively undersampled data for R=5 and R=7 were acquired from 2 volunteers. Bland-Altman analyses were used to assess the agreement between the accelerated and reference T 1ρ measurements. The reconstruction performance was evaluated using the normalized root mean square error and the median of the normalized absolute deviation (MNAD) of the reconstructed T 1ρ -weighted images and the corresponding T 1ρ maps.Results: T 1ρ parameter maps were successfully estimated from T 1ρ -weighted images reconstructed using the proposed method for all accelerations. The accelerated T 1ρ measurements and reference values were in good agreement for R=5.2 (T 1ρ : 40.4±1.4 ms), R=7.7 (T 1ρ : 40.4±2.1 ms), and R=9.7 (T 1ρ : 40.9±2.2 ms) in the Bland-Altman analyses. The T 1ρ parameter maps reconstructed from the prospectively undersampled data also showed promising image quality using the proposed method.
Conclusions:The proposed method achieves the 3D T 1ρ mapping of in vivo knee cartilage in eight minutes using a signal-compensated robust tensor principal component analysis method in image reconstruction.
Denoising diffusion probabilistic models (DDPMs) have been shown to have superior performances in MRI reconstruction. From the perspective of continuous stochastic differential equations (SDEs), the reverse process of DDPM can be seen as maximizing the energy of the reconstructed MR image, leading to SDE sequence divergence. For this reason, a modified high-frequency DDPM model is proposed for MRI reconstruction. From its continuous SDE viewpoint, termed high-frequency space SDE (HFS-SDE), the energy concentrated low-frequency part of the MR image is no longer amplified, and the diffusion process focuses more on acquiring high-frequency prior information. It not only improves the stability of the diffusion model but also provides the possibility of better recovery of high-frequency details. Experiments on the publicly fastMRI dataset show that our proposed HFS-SDE outperforms the DDPM-driven VP-SDE, supervised deep learning methods and traditional parallel imaging methods in terms of stability and reconstruction accuracy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.