Metal Artifact Correction in Industrial CT Images Based on a Dual-Domain Joint Deep Learning Framework

被引:1
作者
Jiang, Shibo [1 ,2 ]
Sun, Yuewen [1 ,2 ]
Xu, Shuo [1 ,2 ]
Zhang, Zehuan [1 ,2 ]
Wu, Zhifang [1 ,2 ]
机构
[1] Tsinghua Univ, Inst Nucl & New Energy Technol, Beijing 100084, Peoples R China
[2] Tsinghua Univ, Beijing Key Lab Nucl Detect Technol, Beijing 100084, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2024年 / 14卷 / 08期
关键词
industrial CT; metal artifact correction; dual-domain; UNet; ResNet; joint framework; REDUCTION;
D O I
10.3390/app14083261
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Industrial computed tomography (CT) images reconstructed directly from projection data using the filtered back projection (FBP) method exhibit strong metal artifacts due to factors such as beam hardening, scatter, statistical noise, and deficiencies in the reconstruction algorithms. Traditional correction approaches, confined to either the projection domain or the image domain, fail to fully utilize the rich information embedded in the data. To leverage information from both domains, we propose a joint deep learning framework that integrates UNet and ResNet architectures for the correction of metal artifacts in CT images. Initially, the UNet network is employed to correct the imperfect projection data (sinograms), the output of which serves as the input for the CT image reconstruction unit. Subsequently, the reconstructed CT images are fed into the ResNet, with both networks undergoing a joint training process to optimize image quality. We take the projection data obtained by analytical simulation as the data set. The resulting optimized industrial CT images show a significant reduction in metal artifacts, with the average Peak Signal-to-Noise Ratio (PSNR) reaching 36.13 and the average Structural Similarity Index (SSIM) achieving 0.953. By conducting simultaneous correction in both the projection and image domains, our method effectively harnesses the complementary information from both, exhibiting a marked improvement in correction results over the deep learning-based single-domain corrections. The generalization capability of our proposed method is further verified in ablation experiments and multi-material phantom CT artifact correction.
引用
收藏
页数:15
相关论文
共 50 条
[21]   U-net based metal segmentation on projection domain for metal artifact reduction in dental CT [J].
Hegazy, Mohamed A. A. ;
Cho, Myung Hye ;
Cho, Min Hyoung ;
Lee, Soo Yeol .
BIOMEDICAL ENGINEERING LETTERS, 2019, 9 (03) :375-385
[22]   Semi-supervised segmentation of metal-artifact contaminated industrial CT images using improved CycleGAN [J].
Jiang, Shi Bo ;
Sun, Yue Wen ;
Xu, Shuo ;
Zhang, Hua Xia ;
Wu, Zhi Fang .
JOURNAL OF X-RAY SCIENCE AND TECHNOLOGY, 2024, 32 (02) :271-283
[23]   DPDudoNet: Deep-Prior Based Dual-Domain Network for Low-Dose Computed Tomography Reconstruction [J].
Komolafe, Temitope Emmanuel ;
Sun, Yuhang ;
Wang, Nizhuan ;
Sun, Kaicong ;
Cao, Guohua ;
Shen, Dinggang .
MACHINE LEARNING FOR MEDICAL IMAGE RECONSTRUCTION (MLMIR 2022), 2022, 13587 :123-132
[24]   A Novel Deep Learning Model Based Cancerous Lung Nodules Severity Grading Framework Using CT Images [J].
Kumar, P. Mohan ;
Jayanthi, V. E. .
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY, 2025, 35 (04)
[25]   Wavelet-Based Iterative Network for Dual-Domain Sparse-View CT Reconstruction Using MRI Priors [J].
Yao, Qiulei ;
Cheng, Ying ;
Chen, Jun ;
Wang, Zhe ;
Cao, Guohua .
MEDICAL IMAGING 2025: PHYSICS OF MEDICAL IMAGING, PT 1, 2025, 13405
[26]   Metal Artefact Correction Algorithm based-on DSAT Technique for CT Images [J].
Osman, N. D. ;
Salikin, M. S. ;
Saripan, M. I. ;
Aziz, M. Z. A. ;
Daud, N. M. .
2014 IEEE CONFERENCE ON BIOMEDICAL ENGINEERING AND SCIENCES (IECBES), 2014, :324-327
[27]   A dual-stream deep convolutional network for reducing metal streak artifacts in CT images [J].
Gjesteby, Lars ;
Shan, Hongming ;
Yang, Qingsong ;
Xi, Yan ;
Jin, Yannan ;
Giantsoudi, Drosoula ;
Paganetti, Harald ;
De Man, Bruno ;
Wang, Ge .
PHYSICS IN MEDICINE AND BIOLOGY, 2019, 64 (23)
[28]   Deep Learning Based Metal Inpainting in the Projection Domain: Initial Results [J].
Gottschalk, Tristan M. ;
Kreher, Bjorn W. ;
Kunze, Holger ;
Maier, Andreas .
MACHINE LEARNING FOR MEDICAL IMAGE RECONSTRUCTION, MLMIR 2019, 2019, 11905 :125-136
[29]   Deep Dual-Domain United Guiding Learning With Global-Local Transformer-Convolution U-Net for LDCT Reconstruction [J].
Wu, Zhan ;
Zhong, Xinyun ;
Lyv, Tianling ;
Wang, Dayang ;
Chen, Ruifeng ;
Yan, Xu ;
Coatrieux, Gouenou ;
Ji, Xu ;
Yu, Hengyong ;
Chen, Yang ;
Mai, Xiaoli .
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2023, 72 :1-15
[30]   A Deep Learning-Based Scatter Correction of Simulated X-ray Images [J].
Lee, Heesin ;
Lee, Joonwhoan .
ELECTRONICS, 2019, 8 (09)