Article
Computer Science, Artificial Intelligence
Jose J. Calvino, Elena Fernandez, Miguel Lopez-Haro, Juan M. Munoz-Ocana, Antonio M. Rodriguez-Chia
Summary: This paper introduces an l(1)-norm model based on Total Variation Minimization for tomographic reconstruction. The reconstructions produced by this model are more accurate than classical reconstruction models based on the l(2)-norm. The model can be linearized and solved using linear programming, and the dimension of the formulation can be further reduced by exploiting complementary slackness conditions.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Interdisciplinary Applications
Jianfeng Guo, C. Ross Schmidtlein, Andrzej Krol, Si Li, Yizun Lin, Sangtae Ahn, Charles Stearns, Yuesheng Xu
Summary: We investigated the imaging performance of a fast convergent ordered-subsets algorithm with subiteration-dependent preconditioners (SDPs) for positron emission tomography (PET) image reconstruction. The study showed that the SDP-BSREM algorithms significantly improve the convergence rate compared to conventional BSREM and a vendor's implementation as Q.Clear, providing faster convergence to the same objective function value.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Mathematics
Omar M. Abou Al-Ola, Ryosuke Kasai, Yusaku Yamaguchi, Takeshi Kojima, Tetsuya Yoshinaga
Summary: This article introduces a new algorithm that combines the advantages of different iterative schemes by combining ordered-subsets EM and MART with weighted geometric or hybrid means, achieving a decrease in the objective function with each iteration and outperforming OS-EM and OS-MART alone. The algorithm shows excellent performance in image reconstruction experiments, especially in dealing with noise issues.
Article
Computer Science, Interdisciplinary Applications
Changyu Chen, Yuxiang Xing, Hewei Gao, Li Zhang, Zhiqiang Chen
Summary: This paper proposes a self-augmented multi-stage deep-learning network (Sam's Net) for end-to-end reconstruction of limited angle CT. By integrating alternating minimization technique and multi-stage self-constraints, Sam's Net addresses the structural distortions and artifacts in limited angle reconstruction. Experimental results demonstrate that Sam's Net significantly improves reconstruction quality and exhibits high stability and robustness under different configurations.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Mineka Sato, Yasutaka Ichikawa, Kensuke Domae, Kazuya Yoshikawa, Yoshinori Kanii, Akio Yamazaki, Naoki Nagasawa, Motonori Nagata, Masaki Ishida, Hajime Sakuma
Summary: Compared to hybrid iterative reconstruction (IR), deep learning image reconstruction (DLIR) improves vessel conspicuity, contrast-to-noise ratio (CNR), and lesion conspicuity of virtual monochromatic and iodine density images in abdominal contrast-enhanced dual-energy computed tomography (DECT).
EUROPEAN RADIOLOGY
(2022)
Review
Radiology, Nuclear Medicine & Medical Imaging
Norbert J. Pelc, David A. Chesler
Summary: This paper aims to describe the early introduction of CT imaging at Massachusetts General Hospital (MGH), with the installation of the first CT scanner in 1973, and the preceding CT research work and related accomplishments.
Article
Engineering, Multidisciplinary
M. R. Qader
Summary: This article introduces a new technique for the maintenance scheduling of a hydrothermal power system, analyzing two coordination schemes and proposing two variants of the Lagrangian relaxation method and augmented relaxation method. Through experimental verification, these methods can effectively achieve convergence and minimum total cost, providing methods for the arrangement of hydro maintenance schedule.
ALEXANDRIA ENGINEERING JOURNAL
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Le Cao, Xiang Liu, Tingting Qu, Yannan Cheng, Jianying Li, Yanan Li, Lihong Chen, Xinyi Niu, Qian Tian, Jianxin Guo
Summary: This study evaluated the use of thin slices and deep learning image reconstruction (DLIR) in contrast-enhanced abdominal CT. The results showed that DLIR significantly reduced image noise and improved image quality and diagnostic confidence.
EUROPEAN RADIOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Bingqian Chu, Lu Gan, Yi Shen, Jian Song, Ling Liu, Jianying Li, Bin Liu
Summary: This study compared the performance of deep learning image reconstruction (DLIR) and adaptive statistical iterative reconstruction-Veo (ASIR-V) in improving image quality and diagnostic performance using virtual monochromatic spectral images in abdominal dual-energy computed tomography (DECT). The results showed that DLIR-M and DLIR-H had better image quality and higher diagnostic confidence compared to ASIR-V40%. DLIR-H performed the best among all the reconstruction methods. The study also found that DECT with DLIR-H further reduced image noise and improved image quality compared to ASIR-V40%.
JOURNAL OF DIGITAL IMAGING
(2023)
Article
Engineering, Biomedical
Genwei Ma, Xing Zhao, Yining Zhu, Huitao Zhang
Summary: This study proposes a novel lightweight block reconstruction network (LBRN) to address the issue of memory space in learning-based CT reconstruction. The network transforms the reconstruction operator into a deep neural network and consists of two main modules for filtering and back-projection. The first module decouples the relationship between the reconstructed image and the projection data, enabling the following block back-projection module to use a simplified block reconstruction strategy. The approach is trained end-to-end, working directly from raw projection data without relying on initial images.
PHYSICS IN MEDICINE AND BIOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Maocan Song, Lin Cheng
Summary: With the increasing demand for travel and limited transportation resources, traffic congestion remains a challenging problem. This study considers the variability of travel times in vehicle routing optimization by utilizing historical travel time data. A mean-standard deviation based vehicle routing model is developed and solved using an augmented Lagrangian relaxation approach. The proposed method effectively reduces the relative gap between the lower and upper bounds in the solving procedure.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Multidisciplinary Sciences
Shiwo Deng, Yining Zhu, Huitao Zhang, Qian Wang, Peiping Zhu, Kai Zhang, Peng Zhang
Summary: Material decomposition is an important application of computer tomography. A new method called PAMD-SART is developed for incorporating both phase and absorption information into the MD process, leading to superior noise suppression and accurate decomposition compared to traditional two-step methods. Numerical simulations and experiments show that PAMD-SART outperforms classical MD methods in terms of quantitative accuracy of material equivalent atomic number.
Article
Computer Science, Interdisciplinary Applications
Xi Tao, Yongbo Wang, Liyan Lin, Zixuan Hong, Jianhua Ma
Summary: The study trains DNN to reconstruct CT images from VVBP-Tensor, providing lossless information and preserving fine details. The learning strategy can be seen as a generalization of conventional methods and potentially inspire algorithm development. The VVBP-Tensor domain learning framework shows significant improvement over traditional image and projection-based learning frameworks.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2021)
Article
Optics
Zhen Guo, Jung Ki Song, George Barbastathis, Michael E. Glinsky, Courtenay T. Vaughan, Kurt W. Larson, Bradley K. Alpert, Zachary H. Levine
Summary: Researchers have developed a Physics-assisted Generative Adversarial Network (PGAN) algorithm for reconstruction in X-ray tomography. Compared to previous methods, PGAN combines known physics and learned prior to reduce photon requirement and achieve a given error rate by reducing projection angles.
Article
Computer Science, Artificial Intelligence
Maocan Song, Lin Cheng, Bin Lu
Summary: This study focuses on solving the multi-compartment vehicle routing problem and proposes two new formulations. By applying Lagrangian relaxation and decomposition techniques, high-quality feasible solutions are obtained.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)