1
|
Zhao F, Liu M, Xiang M, Li D, Jiang X, Jin X, Lin C, Wang R. Unsupervised and Self-supervised Learning in Low-Dose Computed Tomography Denoising: Insights from Training Strategies. JOURNAL OF IMAGING INFORMATICS IN MEDICINE 2024:10.1007/s10278-024-01213-8. [PMID: 39231886 DOI: 10.1007/s10278-024-01213-8] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/13/2024] [Revised: 06/30/2024] [Accepted: 07/01/2024] [Indexed: 09/06/2024]
Abstract
In recent years, X-ray low-dose computed tomography (LDCT) has garnered widespread attention due to its significant reduction in the risk of patient radiation exposure. However, LDCT images often contain a substantial amount of noises, adversely affecting diagnostic quality. To mitigate this, a plethora of LDCT denoising methods have been proposed. Among them, deep learning (DL) approaches have emerged as the most effective, due to their robust feature extraction capabilities. Yet, the prevalent use of supervised training paradigms is often impractical due to the challenges in acquiring low-dose and normal-dose CT pairs in clinical settings. Consequently, unsupervised and self-supervised deep learning methods have been introduced for LDCT denoising, showing considerable potential for clinical applications. These methods' efficacy hinges on training strategies. Notably, there appears to be no comprehensive reviews of these strategies. Our review aims to address this gap, offering insights and guidance for researchers and practitioners. Based on training strategies, we categorize the LDCT methods into six groups: (i) cycle consistency-based, (ii) score matching-based, (iii) statistical characteristics of noise-based, (iv) similarity-based, (v) LDCT synthesis model-based, and (vi) hybrid methods. For each category, we delve into the theoretical underpinnings, training strategies, strengths, and limitations. In addition, we also summarize the open source codes of the reviewed methods. Finally, the review concludes with a discussion on open issues and future research directions.
Collapse
Affiliation(s)
- Feixiang Zhao
- School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China
- College of Nuclear Technology and Automation Engineering, Chengdu University of Technology, 1 East Third Road, Chengdu, 610059, Sichuan, China
| | - Mingzhe Liu
- School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China
- College of Computer Science and Cyber Security, Chengdu University of Technology, 1 East Third Road, Chengdu, 610059, Sichuan, China
| | - Mingrong Xiang
- School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China.
- School of Information Technology, Deakin University, Melbourne Burwood Campus, 221 Burwood Hwy, Melbourne, 3125, Victoria, Australia.
| | - Dongfen Li
- College of Computer Science and Cyber Security, Chengdu University of Technology, 1 East Third Road, Chengdu, 610059, Sichuan, China
| | - Xin Jiang
- School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China
| | - Xiance Jin
- Department of Radiotherapy Center, The first Affiliated Hospital of Wenzhou Medical University, Ouhai District, Wenzhou, 325000, Zhejiang, China
| | - Cai Lin
- Department of Burn, Wound Repair and Regenerative Medicine Center, The first Affiliated Hospital of Wenzhou Medical University, Ouhai District, Wenzhou, 325000, Zhejiang, China
| | - Ruili Wang
- School of Data Science and Artificial Intelligence, Wenzhou University of Technology, Ouhai District, Wenzhou, 325000, Zhejiang, China
- School of Mathematical and Computational Science, Massey University, SH17, Albany, 0632, Auckland, New Zealand
| |
Collapse
|
2
|
Ni G, Wu R, Zheng F, Li M, Huang S, Ge X, Liu L, Liu Y. Toward Ground-Truth Optical Coherence Tomography via Three-Dimensional Unsupervised Deep Learning Processing and Data. IEEE TRANSACTIONS ON MEDICAL IMAGING 2024; 43:2395-2407. [PMID: 38324426 DOI: 10.1109/tmi.2024.3363416] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 02/09/2024]
Abstract
Optical coherence tomography (OCT) can perform non-invasive high-resolution three-dimensional (3D) imaging and has been widely used in biomedical fields, while it is inevitably affected by coherence speckle noise which degrades OCT imaging performance and restricts its applications. Here we present a novel speckle-free OCT imaging strategy, named toward-ground-truth OCT ( t GT-OCT), that utilizes unsupervised 3D deep-learning processing and leverages OCT 3D imaging features to achieve speckle-free OCT imaging. Specifically, our proposed t GT-OCT utilizes an unsupervised 3D-convolution deep-learning network trained using random 3D volumetric data to distinguish and separate speckle from real structures in 3D imaging volumetric space; moreover, t GT-OCT effectively further reduces speckle noise and reveals structures that would otherwise be obscured by speckle noise while preserving spatial resolution. Results derived from different samples demonstrated the high-quality speckle-free 3D imaging performance of t GT-OCT and its advancement beyond the previous state-of-the-art. The code is available online: https://github.com/Voluntino/tGT-OCT.
Collapse
|
3
|
An R, Chen K, Li H. Self-supervised dual-domain balanced dropblock-network for low-dose CT denoising. Phys Med Biol 2024; 69:075026. [PMID: 38359449 DOI: 10.1088/1361-6560/ad29ba] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/06/2023] [Accepted: 02/15/2024] [Indexed: 02/17/2024]
Abstract
Objective.Self-supervised learning methods have been successfully applied for low-dose computed tomography (LDCT) denoising, with the advantage of not requiring labeled data. Conventional self-supervised methods operate only in the image domain, ignoring valuable priors in the sinogram domain. Recently proposed dual-domain methods address this limitation but encounter issues with blurring artifacts in the reconstructed image due to the inhomogeneous distribution of noise levels in low-dose sinograms.Approach.To tackle this challenge, this paper proposes SDBDNet, an end-to-end dual-domain self-supervised method for LDCT denoising. With the network designed based on the properties of inhomogeneous noise in low-dose sinograms and the principle of moderate sinogram-domain denoising, SDBDNet achieves effective denoising in dual domains without introducing blurring artifacts. Specifically, we split the sinogram into two subsets based on the positions of detector cells to generate paired training data with high similarity and independent noise. These sub-sinograms are then restored to their original size using 1D interpolation and learning-based correction. To achieve adaptive and moderate smoothing in the sinogram domain, we integrate Dropblock, a type of convolution layer with regularization, into SDBDNet, and set a weighted average between the denoised sinograms and their noisy counterparts, leading to a well-balanced dual-domain approach.Main results.Numerical experiments show that our method outperforms popular non-learning and self-supervised learning methods, demonstrating its effectiveness and superior performance.Significance.While introducing a novel high-performance dual-domain self-supervised LDCT denoising method, this paper also emphasizes and verifies the importance of appropriate sinogram-domain denoising in dual-domain methods, which might inspire future work.
Collapse
Affiliation(s)
- Ran An
- School of Mathematical Sciences, Capital Normal University, Beijing, 100048, People's Republic of China
- Centre for Mathematical Imaging Techniques, University of Liverpool, Liverpool, L69 7ZL, United Kingdom
| | - Ke Chen
- Department of Mathematics and Statistics, University of Strathclyde, Glasgow, G1 1XQ, United Kingdom
| | - Hongwei Li
- School of Mathematical Sciences, Capital Normal University, Beijing, 100048, People's Republic of China
| |
Collapse
|
4
|
Bousse A, Kandarpa VSS, Rit S, Perelli A, Li M, Wang G, Zhou J, Wang G. Systematic Review on Learning-based Spectral CT. IEEE TRANSACTIONS ON RADIATION AND PLASMA MEDICAL SCIENCES 2024; 8:113-137. [PMID: 38476981 PMCID: PMC10927029 DOI: 10.1109/trpms.2023.3314131] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/14/2024]
Abstract
Spectral computed tomography (CT) has recently emerged as an advanced version of medical CT and significantly improves conventional (single-energy) CT. Spectral CT has two main forms: dual-energy computed tomography (DECT) and photon-counting computed tomography (PCCT), which offer image improvement, material decomposition, and feature quantification relative to conventional CT. However, the inherent challenges of spectral CT, evidenced by data and image artifacts, remain a bottleneck for clinical applications. To address these problems, machine learning techniques have been widely applied to spectral CT. In this review, we present the state-of-the-art data-driven techniques for spectral CT.
Collapse
Affiliation(s)
- Alexandre Bousse
- LaTIM, Inserm UMR 1101, Université de Bretagne Occidentale, 29238 Brest, France
| | | | - Simon Rit
- Univ Lyon, INSA-Lyon, Université Claude Bernard Lyon 1, UJM-Saint Étienne, CNRS, Inserm, CREATIS UMR 5220, U1294, F-69373, Lyon, France
| | - Alessandro Perelli
- Department of Biomedical Engineering, School of Science and Engineering, University of Dundee, DD1 4HN, UK
| | - Mengzhou Li
- Biomedical Imaging Center, Rensselaer Polytechnic Institute, Troy, New York, USA
| | - Guobao Wang
- Department of Radiology, University of California Davis Health, Sacramento, USA
| | - Jian Zhou
- CTIQ, Canon Medical Research USA, Inc., Vernon Hills, 60061, USA
| | - Ge Wang
- Biomedical Imaging Center, Rensselaer Polytechnic Institute, Troy, New York, USA
| |
Collapse
|
5
|
Kim W, Lee J, Choi JH. An unsupervised two-step training framework for low-dose computed tomography denoising. Med Phys 2024; 51:1127-1144. [PMID: 37432026 DOI: 10.1002/mp.16628] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/03/2023] [Revised: 06/25/2023] [Accepted: 06/25/2023] [Indexed: 07/12/2023] Open
Abstract
BACKGROUND Although low-dose computed tomography (CT) imaging has been more widely adopted in clinical practice to reduce radiation exposure to patients, the reconstructed CT images tend to have more noise, which impedes accurate diagnosis. Recently, deep neural networks using convolutional neural networks to reduce noise in the reconstructed low-dose CT images have shown considerable improvement. However, they need a large number of paired normal- and low-dose CT images to fully train the network via supervised learning methods. PURPOSE To propose an unsupervised two-step training framework for image denoising that uses low-dose CT images of one dataset and unpaired high-dose CT images from another dataset. METHODS Our proposed framework trains the denoising network in two steps. In the first training step, we train the network using 3D volumes of CT images and predict the center CT slice from them. This pre-trained network is used in the second training step to train the denoising network and is combined with the memory-efficient denoising generative adversarial network (DenoisingGAN), which further enhances both objective and perceptual quality. RESULTS The experimental results on phantom and clinical datasets show superior performance over the existing traditional machine learning and self-supervised deep learning methods, and the results are comparable to the fully supervised learning methods. CONCLUSIONS We proposed a new unsupervised learning framework for low-dose CT denoising, convincingly improving noisy CT images from both objective and perceptual quality perspectives. Because our denoising framework does not require physics-based noise models or system-dependent assumptions, our proposed method can be easily reproduced; consequently, it can also be generally applicable to various CT scanners or dose levels.
Collapse
Affiliation(s)
- Wonjin Kim
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| | - Jaayeon Lee
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| | - Jang-Hwan Choi
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| |
Collapse
|
6
|
Sadia RT, Chen J, Zhang J. CT image denoising methods for image quality improvement and radiation dose reduction. J Appl Clin Med Phys 2024; 25:e14270. [PMID: 38240466 PMCID: PMC10860577 DOI: 10.1002/acm2.14270] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/18/2023] [Revised: 12/15/2023] [Accepted: 12/28/2023] [Indexed: 02/13/2024] Open
Abstract
With the ever-increasing use of computed tomography (CT), concerns about its radiation dose have become a significant public issue. To address the need for radiation dose reduction, CT denoising methods have been widely investigated and applied in low-dose CT images. Numerous noise reduction algorithms have emerged, such as iterative reconstruction and most recently, deep learning (DL)-based approaches. Given the rapid advancements in Artificial Intelligence techniques, we recognize the need for a comprehensive review that emphasizes the most recently developed methods. Hence, we have performed a thorough analysis of existing literature to provide such a review. Beyond directly comparing the performance, we focus on pivotal aspects, including model training, validation, testing, generalizability, vulnerability, and evaluation methods. This review is expected to raise awareness of the various facets involved in CT image denoising and the specific challenges in developing DL-based models.
Collapse
Affiliation(s)
- Rabeya Tus Sadia
- Department of Computer ScienceUniversity of KentuckyLexingtonKentuckyUSA
| | - Jin Chen
- Department of Medicine‐NephrologyUniversity of Alabama at BirminghamBirminghamAlabamaUSA
| | - Jie Zhang
- Department of RadiologyUniversity of KentuckyLexingtonKentuckyUSA
| |
Collapse
|
7
|
Niu C, Li M, Fan F, Wu W, Guo X, Lyu Q, Wang G. Noise Suppression With Similarity-Based Self-Supervised Deep Learning. IEEE TRANSACTIONS ON MEDICAL IMAGING 2023; 42:1590-1602. [PMID: 37015446 PMCID: PMC10288330 DOI: 10.1109/tmi.2022.3231428] [Citation(s) in RCA: 7] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/02/2023]
Abstract
Image denoising is a prerequisite for downstream tasks in many fields. Low-dose and photon-counting computed tomography (CT) denoising can optimize diagnostic performance at minimized radiation dose. Supervised deep denoising methods are popular but require paired clean or noisy samples that are often unavailable in practice. Limited by the independent noise assumption, current self-supervised denoising methods cannot process correlated noises as in CT images. Here we propose the first-of-its-kind similarity-based self-supervised deep denoising approach, referred to as Noise2Sim, that works in a nonlocal and nonlinear fashion to suppress not only independent but also correlated noises. Theoretically, Noise2Sim is asymptotically equivalent to supervised learning methods under mild conditions. Experimentally, Nosie2Sim recovers intrinsic features from noisy low-dose CT and photon-counting CT images as effectively as or even better than supervised learning methods on practical datasets visually, quantitatively and statistically. Noise2Sim is a general self-supervised denoising approach and has great potential in diverse applications.
Collapse
|
8
|
Hong Z, Zeng D, Tao X, Ma J. Learning CT projection denoising from adjacent views. Med Phys 2023; 50:1367-1377. [PMID: 36414024 DOI: 10.1002/mp.16115] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/03/2022] [Revised: 10/06/2022] [Accepted: 10/29/2022] [Indexed: 11/24/2022] Open
Abstract
BACKGROUND Many learning-based low-dose (LD) computed tomography (CT) imaging methods require large paired full- and low-dose datasets for training, which are usually unavailable in clinic. Whereas models trained on simulated data often face the generalization problem on real clinical data. PURPOSE To develop an unsupervised learning technique to acquire clean CT projection from its adjacent LD projections. METHODS Given a sequential LD projection set, the method extracts out the middle projection as the target and treats the rest ones as the input. The model is trained with the mean absolute error with proposed inter-view gradient constraint term, which helps to suppress outliers and preserve edges in the denoised projection. The simulated low-dose CT grand challenge dataset and a real physical torso phantom dataset were employed for experiment. The peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) were calculated for quantitative evaluation. RESULTS In experiments with both the simulated and real datasets, visual comparisons reveal that the proposed method obtained images superior to unsupervised and supervised methods working in both image and projection domain. For numerical comparison, our method obtains larger SSIMs than other unsupervised methods at quarter and eighth dose levels. As for PSNR, our method obtains larger value at eighth dose whereas smaller value at quarter dose. The supervised models obtain better numerical results than all unsupervised models on simulated datasets. CONCLUSION The proposed method can reduce the noise in CT projections effectively, making it an attractive tool for practical LDCT pre-processing.
Collapse
Affiliation(s)
- Zixuan Hong
- School of Biomedical Engineering, Southern Medical University, Guangdong, China
- Guangdong Artificial Intelligence and Digital Economy Laboratory (Guangzhou), Guangdong, China
| | - Dong Zeng
- School of Biomedical Engineering, Southern Medical University, Guangdong, China
- Guangdong Artificial Intelligence and Digital Economy Laboratory (Guangzhou), Guangdong, China
| | - Xi Tao
- School of Biomedical Engineering, Southern Medical University, Guangdong, China
- Guangdong Artificial Intelligence and Digital Economy Laboratory (Guangzhou), Guangdong, China
| | - Jianhua Ma
- School of Biomedical Engineering, Southern Medical University, Guangdong, China
- Guangdong Artificial Intelligence and Digital Economy Laboratory (Guangzhou), Guangdong, China
| |
Collapse
|
9
|
Ren J, Liang N, Yu X, Wang Y, Cai A, Li L, Yan B. Projection domain processing for low-dose CT reconstruction based on subspace identification. JOURNAL OF X-RAY SCIENCE AND TECHNOLOGY 2023; 31:63-84. [PMID: 36314189 DOI: 10.3233/xst-221262] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/16/2023]
Abstract
PURPOSE Low-dose computed tomography (LDCT) has promising potential for dose reduction in medical applications, while suffering from low image quality caused by noise. Therefore, it is in urgent need for developing new algorithms to obtain high-quality images for LDCT. METHODS This study tries to exploit the sparse and low-rank properties of images and proposes a new algorithm based on subspace identification. The collection of transmission data is sparsely represented by singular value decomposition and the eigen-images are then denoised by block-matching frames. Then, the projection is regularized by the correlation information under the frame of prior image compressed sensing (PICCS). With the application of a typical analytical algorithm on the processed projection, the target images are obtained. Both numerical simulations and real data verifications are carried out to test the proposed algorithm. The numerical simulations data is obtained based on real clinical scanning three-dimensional data and the real data is obtained by scanning experimental head phantom. RESULTS In simulation experiment, using new algorithm boots the means of PSNR and SSIM by 1 dB and 0.05, respectively, compared with BM3D under the Gaussian noise with variance 0.04. Meanwhile, on the real data, the proposed algorithm exhibits superiority over compared algorithms in terms of noise suppression, detail preservation and computational overhead. The means of PSNR and SSIM are improved by 1.84 dB and 0.1, respectively, compared with BM3D under the Gaussian noise with variance 0.04. CONCLUSION This study demonstrates the feasibility and advantages of a new algorithm based on subspace identification for LDCT. It exploits the similarity among three-dimensional data to improve the image quality in a concise way and shows a promising potential on future clinical diagnosis.
Collapse
Affiliation(s)
- Junru Ren
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Ningning Liang
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Xiaohuan Yu
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Yizhong Wang
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Ailong Cai
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Lei Li
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| | - Bin Yan
- Henan Key Laboratory of Imaging and Intelligent Processing, PLA Strategy Support Force Information Engineering University, Zhengzhou, Henan, China
| |
Collapse
|
10
|
Kim W, Lee J, Kang M, Kim JS, Choi JH. Wavelet subband-specific learning for low-dose computed tomography denoising. PLoS One 2022; 17:e0274308. [PMID: 36084002 PMCID: PMC9462582 DOI: 10.1371/journal.pone.0274308] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/11/2022] [Accepted: 08/25/2022] [Indexed: 11/19/2022] Open
Abstract
Deep neural networks have shown great improvements in low-dose computed tomography (CT) denoising. Early algorithms were primarily optimized to obtain an accurate image with low distortion between the denoised image and reference full-dose image at the cost of yielding an overly smoothed unrealistic CT image. Recent research has sought to preserve the fine details of denoised images with high perceptual quality, which has been accompanied by a decrease in objective quality due to a trade-off between perceptual quality and distortion. We pursue a network that can generate accurate and realistic CT images with high objective and perceptual quality within one network, achieving a better perception-distortion trade-off. To achieve this goal, we propose a stationary wavelet transform-assisted network employing the characteristics of high- and low-frequency domains of the wavelet transform and frequency subband-specific losses defined in the wavelet domain. We first introduce a stationary wavelet transform for the network training procedure. Then, we train the network using objective loss functions defined for high- and low-frequency domains to enhance the objective quality of the denoised CT image. With this network design, we train the network again after replacing the objective loss functions with perceptual loss functions in high- and low-frequency domains. As a result, we acquired denoised CT images with high perceptual quality using this strategy while minimizing the objective quality loss. We evaluated our algorithms on the phantom and clinical images, and the quantitative and qualitative results indicate that ours outperform the existing state-of-the-art algorithms in terms of objective and perceptual quality.
Collapse
Affiliation(s)
- Wonjin Kim
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| | - Jaayeon Lee
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| | - Mihyun Kang
- Department of Cyber Security, Ewha Womans University, Seoul, Republic of Korea
| | - Jin Sung Kim
- Department of Radiation Oncology, Yonsei University College of Medicine, Seoul, Republic of Korea
| | - Jang-Hwan Choi
- Division of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, Republic of Korea
| |
Collapse
|
11
|
Zhang Z, Yu S, Qin W, Liang X, Xie Y, Cao G. Self-supervised CT super-resolution with hybrid model. Comput Biol Med 2021; 138:104775. [PMID: 34666243 DOI: 10.1016/j.compbiomed.2021.104775] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/28/2021] [Revised: 08/14/2021] [Accepted: 08/17/2021] [Indexed: 12/19/2022]
Abstract
Software-based methods can improve CT spatial resolution without changing the hardware of the scanner or increasing the radiation dose to the object. In this work, we aim to develop a deep learning (DL) based CT super-resolution (SR) method that can reconstruct low-resolution (LR) sinograms into high-resolution (HR) CT images. We mathematically analyzed imaging processes in the CT SR imaging problem and synergistically integrated the SR model in the sinogram domain and the deblur model in the image domain into a hybrid model (SADIR). SADIR incorporates the CT domain knowledge and is unrolled into a DL network (SADIR-Net). The SADIR-Net is a self-supervised network, which can be trained and tested with a single sinogram. SADIR-Net was evaluated through SR CT imaging of a Catphan700 physical phantom and a real porcine phantom, and its performance was compared to the other state-of-the-art (SotA) DL-based CT SR methods. On both phantoms, SADIR-Net obtains the highest information fidelity criterion (IFC), structure similarity index (SSIM), and lowest root-mean-square-error (RMSE). As to the modulation transfer function (MTF), SADIR-Net also obtains the best result and improves the MTF50% by 69.2% and MTF10% by 69.5% compared with FBP. Alternatively, the spatial resolutions at MTF50% and MTF10% from SADIR-Net can reach 91.3% and 89.3% of the counterparts reconstructed from the HR sinogram with FBP. The results show that SADIR-Net can provide performance comparable to the other SotA methods for CT SR reconstruction, especially in the case of extremely limited training data or even no data at all. Thus, the SADIR method could find use in improving CT resolution without changing the hardware of the scanner or increasing the radiation dose to the object.
Collapse
Affiliation(s)
- Zhicheng Zhang
- Department of Radiation Oncology, Stanford University, Stanford, 94305-5847, CA, USA; Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China
| | - Shaode Yu
- College of Information and Communication Engineering, Communication University of China, Beijing 100024, China
| | - Wenjian Qin
- Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China
| | - Xiaokun Liang
- Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China
| | - Yaoqin Xie
- Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China.
| | - Guohua Cao
- Virginia Polytechnic Institute & State University, Blacksburg, VA 24061, USA.
| |
Collapse
|