Lei Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Lei Wang has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Yuhang Liu
LIST, Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications, Ministry of Education, and Jiangsu Provincial Joint International Research Laboratory of Medical Information Processing, Southeast University, Nanjing, China
6 shared publications - Zhiguo Gui
North University of China
6 shared publications - Yi Liu
North University of China
5 shared publications - Pengcheng Zhang
North University of China
4 shared publications - Huazhong Shu
School of Computer Science and Engineering, Southeast University, Nanjing, China
3 shared publications - Fuzhi Wu
College of Information Science and Technology & Artificial Intelligence, Nanjing Forestry University, No. 159 Longpan Road, Nanjing, 210037, CHINA
2 shared publications
Latest Publications
- 24 November 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 301
MRED-Net: A pure tokens-to-token visual mamba-based residual encoder-decoder network for low-dose CT denoising
- 24 November 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 301
MRED-Net: A pure tokens-to-token visual mamba-based residual encoder-decoder network for low-dose CT denoising
- 29 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 72 (7) , 2110-2121
3-D Computed Laminography Based on a Sequential Regularization
Accurate reconstruction of computed laminography (CL) remains challenging due to incomplete projections causing inter-slice aliasing and blurring. In this paper, we propose a novel 3-D CL reconstruction model named SART-SR, which extends traditional single-term regularization methods into a sequential regularization (SR) framework specifically designed for anisotropic CL data. Guided by the theory of "visible and invisible boundaries", this framework decomposes the regularization process into three directional-aware stages: (1) 1-D directional gradient sparsity terms are first applied in the in-slice to enhance reliable edge structures; (2) mild edge-preserving smoothing is applied along the z-direction to reduce aliasing; and (3) a truncated adaptive-weighted total variation (TAwTV) is used for volumetric consistency and streak artifact suppression. To solve the model efficiently, we develop an alternating minimization algorithm based on the Split-Bregman method and gradient descent. The results on simulated MPCB and flange plate phantoms demonstrate that SART-SR notably outperforms competing iterative methods, including simultaneous algebraic reconstruction technique (SART), in preserving edges, suppressing inter-slice aliasing, and reducing noise. The code is publicly available at https://github.com/YuhangLiu98/SART-SR.
- 29 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Nuclear Science
- Vol. 72 (7) , 2110-2121
3-D Computed Laminography Based on a Sequential Regularization
Accurate reconstruction of computed laminography (CL) remains challenging due to incomplete projections causing inter-slice aliasing and blurring. In this paper, we propose a novel 3-D CL reconstruction model named SART-SR, which extends traditional single-term regularization methods into a sequential regularization (SR) framework specifically designed for anisotropic CL data. Guided by the theory of "visible and invisible boundaries", this framework decomposes the regularization process into three directional-aware stages: (1) 1-D directional gradient sparsity terms are first applied in the in-slice to enhance reliable edge structures; (2) mild edge-preserving smoothing is applied along the z-direction to reduce aliasing; and (3) a truncated adaptive-weighted total variation (TAwTV) is used for volumetric consistency and streak artifact suppression. To solve the model efficiently, we develop an alternating minimization algorithm based on the Split-Bregman method and gradient descent. The results on simulated MPCB and flange plate phantoms demonstrate that SART-SR notably outperforms competing iterative methods, including simultaneous algebraic reconstruction technique (SART), in preserving edges, suppressing inter-slice aliasing, and reducing noise. The code is publicly available at https://github.com/YuhangLiu98/SART-SR.
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
SDCNN: Self-Supervised Disentangled Convolutional Neural Network for Low-Dose CT Denoising
Low-dose computed tomography (LDCT) reduces radiation exposure but suffers from high noise, impacting image quality and diagnostic accuracy. Supervised learning has helped address this challenge but requires numerous paired datasets of LDCT and normal-dose CT (NDCT) images, which limits their clinical practice. This article proposes a novel self-supervised disentangled convolutional neural network (SDCNN) that can directly reconstruct high-quality CT images from LDCT data without the need for a clean reference. Unlike other methods that treat noise as a uniform entity, SDCNN disentangles LDCT images into noise-free images, signal-dependent noise, and signal-independent noise, aligning with the intrinsic principles of low-dose noise generation. To enhance the purity of disentanglement, we introduce the concept of combination and re-disentanglement to establish a training framework based on SDCNN. Additionally, we design self-supervised loss functions, including novel anisotropic total variation (TV) and distance loss functions, to improve the efficiency of the denoising process. The signal-guided attention (SGA) module effectively captures the relationship between signal-dependent noise and the signal across both spatial and channel dimensions. Experiments on clinical and animal data demonstrate that the proposed method performed better than all competing state-of-the-art (SOTA) self-supervised algorithms in noise and artifact removal. For example, compared to the self-supervised algorithms, SDCNN improves the multiscale structural similarity (MSSIM), peak signal-to-noise ratio (PSNR), and feature similarity (FSIM) by at least 2.26%, 1.20 dB, 1.23%, whereas, the gradient magnitude similarity deviation (GMSD) is reduced by at least 1.11% on the Mayo clinical data. The code is available at https://github.com/YuhangLiu98/SDCNN .
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
SDCNN: Self-Supervised Disentangled Convolutional Neural Network for Low-Dose CT Denoising
Low-dose computed tomography (LDCT) reduces radiation exposure but suffers from high noise, impacting image quality and diagnostic accuracy. Supervised learning has helped address this challenge but requires numerous paired datasets of LDCT and normal-dose CT (NDCT) images, which limits their clinical practice. This article proposes a novel self-supervised disentangled convolutional neural network (SDCNN) that can directly reconstruct high-quality CT images from LDCT data without the need for a clean reference. Unlike other methods that treat noise as a uniform entity, SDCNN disentangles LDCT images into noise-free images, signal-dependent noise, and signal-independent noise, aligning with the intrinsic principles of low-dose noise generation. To enhance the purity of disentanglement, we introduce the concept of combination and re-disentanglement to establish a training framework based on SDCNN. Additionally, we design self-supervised loss functions, including novel anisotropic total variation (TV) and distance loss functions, to improve the efficiency of the denoising process. The signal-guided attention (SGA) module effectively captures the relationship between signal-dependent noise and the signal across both spatial and channel dimensions. Experiments on clinical and animal data demonstrate that the proposed method performed better than all competing state-of-the-art (SOTA) self-supervised algorithms in noise and artifact removal. For example, compared to the self-supervised algorithms, SDCNN improves the multiscale structural similarity (MSSIM), peak signal-to-noise ratio (PSNR), and feature similarity (FSIM) by at least 2.26%, 1.20 dB, 1.23%, whereas, the gradient magnitude similarity deviation (GMSD) is reduced by at least 1.11% on the Mayo clinical data. The code is available at https://github.com/YuhangLiu98/SDCNN .
- Top Cited Papers
- 3 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Imaging
- Vol. 9 (25730436) , 83-93
Image Denoising for Low-Dose CT via Convolutional Dictionary Learning and Neural Network
Removing noise and artifacts from low-dose computed tomography (LDCT) is a challenging task, and most existing image-based algorithms tend to blur the results. To improve the resolution of denoising results, we combine convolutional dictionary learning and convolutional neural network (CNN), and propose a transfer learning densely connected convolutional dictionary learning (TLD-CDL) framework. In detail, we first introduce the dense connections and multi-scale Inception structure to the network, and train the pre-model on the natural image dataset, then fit the model to the post-processing of LDCT images in the way of transfer learning. In addition, considering that a single pixel-level loss is difficult to achieve satisfactory results both in the index and visual perception, we use the compound loss function of L1 loss and SSIM loss to guide the training. The experimental result shows that TLD-CDL has a good balance between noise reduction and the preservation of details, and acquires inspiring effectiveness in terms of qualitative and quantitative perspective.
- 3 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Imaging
- Vol. 9 (25730436) , 83-93
Top Cited PapersImage Denoising for Low-Dose CT via Convolutional Dictionary Learning and Neural Network
Removing noise and artifacts from low-dose computed tomography (LDCT) is a challenging task, and most existing image-based algorithms tend to blur the results. To improve the resolution of denoising results, we combine convolutional dictionary learning and convolutional neural network (CNN), and propose a transfer learning densely connected convolutional dictionary learning (TLD-CDL) framework. In detail, we first introduce the dense connections and multi-scale Inception structure to the network, and train the pre-model on the natural image dataset, then fit the model to the post-processing of LDCT images in the way of transfer learning. In addition, considering that a single pixel-level loss is difficult to achieve satisfactory results both in the index and visual perception, we use the compound loss function of L1 loss and SSIM loss to guide the training. The experimental result shows that TLD-CDL has a good balance between noise reduction and the preservation of details, and acquires inspiring effectiveness in terms of qualitative and quantitative perspective.
- Open Access
- 5 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 50114-50124
Image Processing for Low-Dose CT via Novel Anisotropic Fourth-Order Diffusion Model
Low-dose CT images contain severe mottle noise and streak artifacts, which seriously affect the physician’s diagnosis of the disease. Hence, in this paper, we propose a novel anisotropic fourth-order diffusion model for low-dose CT image processing. The proposed diffusion model uses both image gradient magnitude and weighted residual local energy to determine the diffusion coefficient. Gradient magnitude is used to detect the image edges, while the weighted residual local energy preserves textures and details in the image. In addition, the fidelity term is introduced into the diffusion model to avoid excessive smoothing and weaken the blocky effects. Experimental results show that when compared with the anisotropic fourth-order diffusion model, the proposed algorithm protects the texture details and suppresses the blocky effects. In comparison with other state-of-the-art algorithms, the proposed model effectively suppresses mottle noise and streak artifacts while simultaneously improving the low-dose CT image quality.
- 5 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 50114-50124
Open AccessImage Processing for Low-Dose CT via Novel Anisotropic Fourth-Order Diffusion Model
Low-dose CT images contain severe mottle noise and streak artifacts, which seriously affect the physician’s diagnosis of the disease. Hence, in this paper, we propose a novel anisotropic fourth-order diffusion model for low-dose CT image processing. The proposed diffusion model uses both image gradient magnitude and weighted residual local energy to determine the diffusion coefficient. Gradient magnitude is used to detect the image edges, while the weighted residual local energy preserves textures and details in the image. In addition, the fidelity term is introduced into the diffusion model to avoid excessive smoothing and weaken the blocky effects. Experimental results show that when compared with the anisotropic fourth-order diffusion model, the proposed algorithm protects the texture details and suppresses the blocky effects. In comparison with other state-of-the-art algorithms, the proposed model effectively suppresses mottle noise and streak artifacts while simultaneously improving the low-dose CT image quality.
- 3 April 2022
- Published by SAGE Publications in Journal of X-Ray Science and Technology
- Vol. 30 (4) , 709-724
Spark plug defects detection based on improved Faster-RCNN algorithm
The objective of this study is to apply an improved Faster-RCNN model in order to solve the problems of low detection accuracy and slow detection speed in spark plug defect detection. In detail, an attention module based symmetrical convolutional network (ASCN) is designed as the backbone to extract multi-scale features. Then, a multi-scale region generation network (MRPN), in which InceptionV2 is used to achieve sliding windows of different scales instead of a single sliding window, is proposed and tested. Additionally, a dataset of X-ray spark plug images is established, which contains 1,402 images. These images are divided into two subsets with a ratio of 4:1 for training and testing the improved Faster-RCNN model, respectively. The proposed model is transferred and learned on the pre-training model of MS COCO dataset. In the test experiments, the proposed method achieves an average accuracy of 89% and a recall of 97%. Compared with other Faster-RCNN models, YOLOv3, SSD and RetinaNet, our proposed new method improves the average accuracy by more than 6% and the recall by more than 2%. Furthermore, the new method can detect at 20fps when the input image size is 1024×1024×3 and can also be used for real-time automatic detection of spark plug defects.
- 3 April 2022
- Published by SAGE Publications in Journal of X-Ray Science and Technology
- Vol. 30 (4) , 709-724
Spark plug defects detection based on improved Faster-RCNN algorithm
The objective of this study is to apply an improved Faster-RCNN model in order to solve the problems of low detection accuracy and slow detection speed in spark plug defect detection. In detail, an attention module based symmetrical convolutional network (ASCN) is designed as the backbone to extract multi-scale features. Then, a multi-scale region generation network (MRPN), in which InceptionV2 is used to achieve sliding windows of different scales instead of a single sliding window, is proposed and tested. Additionally, a dataset of X-ray spark plug images is established, which contains 1,402 images. These images are divided into two subsets with a ratio of 4:1 for training and testing the improved Faster-RCNN model, respectively. The proposed model is transferred and learned on the pre-training model of MS COCO dataset. In the test experiments, the proposed method achieves an average accuracy of 89% and a recall of 97%. Compared with other Faster-RCNN models, YOLOv3, SSD and RetinaNet, our proposed new method improves the average accuracy by more than 6% and the recall by more than 2%. Furthermore, the new method can detect at 20fps when the input image size is 1024×1024×3 and can also be used for real-time automatic detection of spark plug defects.