2-Step Sparse-View CT Reconstruction with a Domain-Specific Perceptual Network

Wei, Haoyu; Schiffers, Florian; Würfl, Tobias; Shen, Daming; Kim, Daniel; Katsaggelos, Aggelos K.; Cossairt, Oliver

doi:10.48550/arxiv.2012.04743

Cited by 3 publications

(8 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We compare our method against [34], which we refer to as score-CT henceforth. We also compare with the best-in-class supervised learning methods, cGAN [12] and SIN-4c-PRN [41]. As a compressed sensing baseline, FISTA-TV [3] was included, along with the analytical reconstruction method, FBP.…”

Section: Datasets and Implementationmentioning

confidence: 99%

Improving Diffusion Models for Inverse Problems using Manifold Constraints

Chung¹,

Sim²,

Ryu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recently, diffusion models have been used to solve various inverse problems in an unsupervised manner with appropriate modifications to the sampling process. However, the current solvers, which recursively apply a reverse diffusion step followed by a measurement consistency step, often produce sub-optimal results. By studying the generative sampling path, here we show that current solvers throw the sample path off the data manifold, and hence the error accumulates. To address this, we propose an additional correction term inspired by the manifold constraint, which can be used synergistically with the previous solvers to make the iterations close to the manifold. The proposed manifold constraint is straightforward to implement within a few lines of code, yet boosts the performance by a surprisingly large margin. With extensive experiments, we show that our method is superior to the previous methods both theoretically and empirically, producing promising results in many applications such as image inpainting, colorization, and sparse-view computed tomography.Nevertheless, for certain problems (e.g. inpainting), currently used algorithms often produce unsatisfactory results when implemented naively (e.g. boundary artifacts, as shown in fig. 1 (b)). The authors in [26] showed that in order to produce high quality reconstructions, one needs to iterate back and forth between the noising and the denoising step at least > 10 times per iteration. These iterations are computationally demanding and should be avoided, considering that diffusion models are slow to sample from even without such iterations.Recently, another type of score-based approach, called Noise2Score [23], was proposed for image denoising without clean references. In contrast to the diffusion models, Noise2Score is computationally efficient as it is a deterministic single-step approach that does not involve any stochastic sampling. Unfortunately, the performance of Noise2Score is inferior to the diffusion models that rely Preprint. Under review.

show abstract

Section: Datasets and Implementationmentioning

confidence: 99%

Improving Diffusion Models for Inverse Problems using Manifold Constraints

Chung¹,

Sim²,

Ryu³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…For MAR experiments, we include another learning-free baseline called linear interpolation (LI, Kalender et al, 1987). (Gilton et al, 2019), and SIN-4c-PRN (Wei et al, 2020) as supervised learning baselines. We follow the settings in Wei et al (2020) and train all methods with 23 projection angles.…”

Section: Standard Techniques In Medical Imagingmentioning

confidence: 99%

“…(Gilton et al, 2019), and SIN-4c-PRN (Wei et al, 2020) as supervised learning baselines. We follow the settings in Wei et al (2020) and train all methods with 23 projection angles. For MAR, we use cGANMAR and SNMAR (Yu et al, 2020) as the baselines.…”

Section: Standard Techniques In Medical Imagingmentioning

confidence: 99%

“…With the rise of machine learning, many methods (Zhu et al, 2018;Mardani et al, 2017;Shen et al, 2019;Würfl et al, 2018;Ghani & Karl, 2018;Wei et al, 2020) have been proposed for medical image reconstruction using a small number of measurements. Most of these methods are supervised learning techniques.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, the images were reconstructed from the overlayed sinogram. Note that we trained the model using 23 projections and tested it on other projection settings to evaluate the generalization.SIN-4c-PRNTo further reduce the artifacts in both sinogram and image space, SIN-4c-PRN(Wei et al, 2020) proposed a two-step sparse-view CT reconstruction model. It involves a sinogram inpainting network (SIN) to generate super-resolved sinograms with different number of projections, and then a post-processing refining network (PRN) to further remove image artifacts.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Solving Inverse Problems in Medical Imaging with Score-Based Generative Models

Song¹,

Shen²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Reconstructing medical images from partial measurements is an important inverse problem in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). Existing solutions based on machine learning typically train a model to directly map measurements to medical images, leveraging a training dataset of paired images and measurements. These measurements are typically synthesized from images using a fixed physical model of the measurement process, which hinders the generalization capability of models to unknown measurement processes. To address this issue, we propose a fully unsupervised technique for inverse problem solving, leveraging the recently introduced score-based generative models. Specifically, we first train a score-based generative model on medical images to capture their prior distribution. Given measurements and a physical model of the measurement process at test time, we introduce a sampling method to reconstruct an image consistent with both the prior and the observed measurements. Our method does not assume a fixed measurement process during training, and can thus be flexibly adapted to different measurement processes at test time. Empirically, we observe comparable or better performance to supervised learning techniques in several medical imaging tasks in CT and MRI, while demonstrating significantly better generalization to unknown measurement processes.

show abstract

DDMM-Synth: A Denoising Diffusion Model for Cross-modal Medical Image Synthesis with Sparse-view Measurement Embedding

Li¹,

Kai²,

Wang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Reducing the radiation dose in computed tomography (CT) is important to mitigate radiation-induced risks. One option is to employ a well-trained model to compensate for incomplete information and map sparse-view measurements to the CT reconstruction. However, reconstruction from sparsely sampled measurements is insufficient to uniquely characterize an object in CT, and a learned prior model may be inadequate for unencountered cases. Medical modal translation from magnetic resonance imaging (MRI) to CT is an alternative but may introduce incorrect information into the synthesized CT images in addition to the fact that there exists no explicit transformation describing their relationship. To address these issues, we propose a novel framework called the denoising diffusion model for medical image synthesis (DDMM-Synth) to close the performance gaps described above. This framework combines an MRI-guided diffusion model with a new CT measurement embedding reverse sampling scheme. Specifically, the null-space content of the one-step denoising result is refined by the MRI-guided data distribution prior, and its range-space component derived from an explicit operator matrix and the sparse-view CT measurements is directly integrated into the inference stage. DDMM-Synth can adjust the projection number of CT a posteriori for a particular clinical application and its modified version can even improve the results significantly for noisy cases. Our results show that DDMM-Synth outperforms other state-of-the-art supervisedlearning-based baselines under fair experimental conditions.

show abstract

2-Step Sparse-View CT Reconstruction with a Domain-Specific Perceptual Network

Cited by 3 publications

References 38 publications

Improving Diffusion Models for Inverse Problems using Manifold Constraints

Improving Diffusion Models for Inverse Problems using Manifold Constraints

Solving Inverse Problems in Medical Imaging with Score-Based Generative Models

DDMM-Synth: A Denoising Diffusion Model for Cross-modal Medical Image Synthesis with Sparse-view Measurement Embedding

Contact Info

Product

Resources

About