Optimizing a Parameterized Plug-and-Play ADMM for Iterative Low-Dose CT Reconstruction

Abstract

Reducing the exposure to X-ray radiation while maintaining a clinically acceptable image quality is desirable in various CT applications. To realize low-dose CT (LdCT) imaging, model-based iterative reconstruction (MBIR) algorithms are widely adopted, but they require proper prior knowledge assumptions in the sinogram and/or image domains and involve tedious manual optimization of multiple parameters. In this work, we propose a deep learning (DL)-based strategy for MBIR to simultaneously address prior knowledge design and MBIR parameter selection in one optimization framework. Specifically, a parameterized plug-and-play alternating direction method of multipliers (3pADMM) is proposed for the general penalized weighted least-squares (PWLS) model, and then, by adopting the basic idea of DL, the parameterized plug-and-play (3p) prior and the related parameters are optimized simultaneously in a single framework using a large number of training data. The main contribution of this work is that the 3p prior and the related parameters in the proposed 3pADMM framework can be supervised and optimized simultaneously to achieve robust LdCT reconstruction performance. Experimental results obtained on clinical patient datasets demonstrate that the proposed method can achieve promising gains over existing algorithms for LdCT image reconstruction in terms of noise-induced artifact suppression and edge detail preservation.

Solving inverse problems using data-driven models

“…This is illustrated in Figure 5.2, and similar networks are also suggested by Vogel and Pock (2017) and Kobler et al (2017), who extend the approach of Hammernik et al (2018) by parametrizing and learning the data discrepancy L. Applications are for inverting the two-dimensional Fourier transform (two-dimensional MRI image reconstruction). See also He et al (2019), who unroll an ADMM scheme with updates in both reconstruction and data spaces and apply that to two-dimensional CT reconstruction. Finally, allowing for some memory in both model parameter and data spaces leads to the learned primal-dual scheme of Adler andÖktem (2018b), which is used for low-dose two-dimensional CT reconstruction.…”

Model Meets Deep Learning in Image Inverse Problems

“…Since then, PnP has been studied extensively with great success. A parameterized PnP-ADMM was proposed in [76]. It used a deep learning-based strategy for model-based iterative reconstruction (MBIR) to simultaneously address the challenges in prior design and MBIR parameter selection.…”

Convolutional Sparse Coding for Compressed Sensing CT Reconstruction

“…Zhu et al proposed a unified image reconstruction framework-automated transform by manifold approximation (AUTOMAP)-which directly learned a potential mapping function from the sensor domain to the image domain [22]. In [23], a parameterized plug-and-play alternating direction method of multipliers (3pADMM) was developed for the PWLS model, and then the prior terms and related parameters were optimized with a neural network with a large quantity of training data. Except for the reconstruction problem, deep learning was also used in low-dose CT (LDCT) denoising.…”