A Noise-Robust Framework for Automatic Segmentation of COVID-19 Pneumonia Lesions From CT Images
Open accessThe DOI matches its Crossref record; 55 of the 57 references it cites were found in a registry or an index.
Abstract
Segmentation of pneumonia lesions from CT scans of COVID-19 patients is important for accurate diagnosis and follow-up. Deep learning has a potential to automate this task but requires a large set of high-quality annotations that are difficult to collect. Learning from noisy training labels that are easier to obtain has a potential to alleviate this problem. To this end, we propose a novel noise-robust framework to learn from noisy labels for the segmentation task. We first introduce a noise-robust Dice loss that is a generalization of Dice loss for segmentation and Mean Absolute Error (MAE) loss for robustness against noise, then propose a novel COVID-19 Pneumonia Lesion segmentation network (COPLE-Net) to better deal with the lesions with various scales and appearances. The noise-robust Dice loss and COPLE-Net are combined with an adaptive self-ensembling framework for training, where an Exponential Moving Average (EMA) of a student model is used as a teacher model that is adaptively updated by suppressing the contribution of the student to EMA when the student has a large training loss. The student model is also adaptive by learning from the teacher only when the teacher outperforms the student. Experimental results showed that: (1) our noise-robust Dice loss outperforms existing noise-robust loss functions, (2) the proposed COPLE-Net achieves higher performance than state-of-the-art image segmentation networks, and (3) our framework with adaptive self-ensembling significantly outperforms a standard training process and surpasses other noise-robust training approaches in the scenario of learning from noisy labels for COVID-19 pneumonia lesion segmentation.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Recently, the Automatic segmentation of COVID-19 Lung Infection from CT Images helps to detect the development of COVID-19 infection by employing several deep learning techniques [309–312].”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [310]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“In addition, in each convolutional block, we use residual connection [27] to facilitate the training and add a Squeeze-and-Excitation (SE) block [40] to calibrate different channels for better performance.”
MethodU-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404
Squeeze-and-Excitation Networks
Hu, J., Shen, L., & Sun, G. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00745 · Cited by 35,667
Focal Loss for Dense Object Detection
Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollar, P. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.324 · Cited by 30,581 · cited 3×
“They were also compared with: standard CE loss L CE , Dice loss L Dice proposed by Milletari et al. [18], L CE + L Dice , L MAE + L Dice , Sørensen-Dice [44], focal CE loss [43] and focal Dice loss [45].”
BackgroundMethodInfluentialA novel coronavirus from patients with pneumonia in China, 2019
Zhu, N., Zhang, D., Wang, W., Li, X., Yang, B., Song, J., Zhao, X., Huang, B., Shi, W., Lu, R., et al. (2020) · New England Journal of Medicine · doi:10.1056/nejmoa2001017 · Cited by 28,244 · cited 2×
“T HE coronavirus disease 2019 (COVID-19) has become a global pandemic since the beginning of 2020 [1]–[3].”
BackgroundEncoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
Chen, L.-C., Zhu, Y., Papandreou, G., Schroff, F., & Adam, H. (2018) · arXiv · doi:10.48550/arxiv.1802.02611 · Cited by 20,063 · cited 2×
“ASPP [35] block is used at the bottleneck to better deal with lesions at multiple scales. bilinear interpolation, where the former reduces the number of feature channels to match that of the output of the bridge layer, and the latter improves the spatial resolution before the concatenation of the…”
MethodEncoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
Chen, L.-C., Zhu, Y., Papandreou, G., Schroff, F., & Adam, H. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01234-2_49 · Cited by 20,054
Squeeze-and-Excitation Networks
Hu, J., Shen, L., Albanie, S., Sun, G., & Wu, E. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2913372 · Cited by 12,157
V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation
Milletari, F., Navab, N., & Ahmadi, S.-A. (2016) · 2016 Fourth International Conference on 3D Vision (3DV) · doi:10.1109/3dv.2016.79 · Cited by 11,931 · cited 10×
“…noisy annotations for training CNNs to segment COVID-19 pneumonia lesions, we propose a novel noise-robust Dice loss function, which is a combination and generalization of MAE loss [14] that is robust against noisy labels and Dice loss [18] that is insensitive to foreground-background imbalance.”
BackgroundMethodInfluentialUNet++: A Nested U-Net Architecture for Medical Image Segmentation
Zhou, Z., Rahman Siddiquee, M. M., Tajbakhsh, N., & Liang, J. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-00889-5_1 · Cited by 9,610
Research works printed but not linked: 1
- [41].Isensee F., Kickingereder P., Wick W., Bendszus M., and Maier-Hein K. H., “No new-net,” in Proc. Int. MICCAI Brainlesion Workshop, 2018, pp. 234–244. [Google Scholar]
Cited by
Items 101–200 of 495
Mardhiah, M., Musgamy, A., & Lubis, M. (2023) · International Journal of Learning Teaching and Educational Research · doi:10.26803/ijlter.22.11.5 · Cited by 38
MultiR-Net: A Novel Joint Learning Network for COVID-19 segmentation and classification
Li, C., Xu, Y.-D., Ding, X., Zhao, J., Du, R.-Q., Wu, L., & Sun, W.-P. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.105340 · Cited by 38
“Wang et al. [26] introduced a noise-robust Dice loss to the network to better deal with the noisy training labels for COVID-19.”
Self-citationSenseCare: A Research Platform for Medical Image Informatics and Interactive 3D VisualizationPreprint
Duan, Q., Wang, G., Wang, R., Fu, C., Li, X., Wang, N., Huang, Y., Huang, X., Song, T., Zhao, L., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.07031 · Cited by 38
Self-citationZhou, M., Xu, Z., Zhou, K. Q., & Tong, R. K. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43895-0_13 · Cited by 37
Self-citationHu, H., Shen, L., Guan, Q., Li, X., Zhou, Q., & Ruan, S. (2021) · Pattern Recognition · doi:10.1016/j.patcog.2021.108452 · Cited by 37
“MosMedData is a dataset of 100 axial CT images from more than 40 patients with COVID-19, including 829 slices with 512x512 size (see [42] for details), and UESTC-COVID-19 contains CT scans (3D volumes) of 50 patients diagnosized with COVID-19 from 10 different hospitals(see [41] for details).”
Meng, Y., Bridge, J., Addison, C. A., Wang, M., Merritt, C., Franks, S., Mackey, M., Messenger, S., Sun, R., FitzMaurice, T. S., et al. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102722 · Cited by 36
“However, some previous methods (Gao et al., 2021; Wu et al., 2021b; Uemura et al., 2021a; Wang et al., 2020d; Liu et al., 2021b; Fan et al., 2020b; Wang et al., 2020e; Zhou et al., 2020; Hou et al., 2021) overlooked the 3D spatial information and developed 2D deep Learning model for the…”
MethodFLANNEL (Focal Loss bAsed Neural Network EnsembLe) for COVID-19 detection
Qiao, Z., Bae, A., Glass, L. M., Xiao, C., & Sun, J. (2020) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocaa280 · Cited by 36
Joshi, A. M., & Nayak, D. R. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3196489 · Cited by 36
“[26] proposed a COVID-19 pneumonia lesion segmentation network (COPLENet) that deals with lesions”
BackgroundCARes‐UNet: Content‐aware residual UNet for lesion segmentation of COVID‐19 from chest CT images
Xu, X., Wen, Y., Zhao, L., Zhang, Y., Zhao, Y., Tang, Z., Yang, Z., & Chen, C. Y. (2021) · Medical Physics · doi:10.1002/mp.15231 · Cited by 35
“Wang et al. 20 proposed a noise-robust framework for COVID-19 lesion segmentation to tackle the inaccurate annotation caused by complex appearances of pneumonia lesions and high inter-and intra-observer variability.”
BackgroundNing, G., Liang, H., Zhang, X., & Liao, H. (2023) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2023.3279114 · Cited by 35
“Restrictions apply. deal with the voids in Doppler images [35]: where p i ∈ R HW and g i ∈ R HW denote the value of each pixel in the segmented mask and Doppler mask.”
BackgroundUnsupervised domain adaptation based COVID-19 CT infection segmentation network
Chen, H., Jiang, Y., Loew, M. H., & Ko, H. (2021) · Applied Intelligence · doi:10.1007/s10489-021-02691-x · Cited by 34
Fan, C., Zeng, Z., Xiao, L., & Qu, X. (2022) · Pattern Recognition · doi:10.1016/j.patcog.2022.108963 · Cited by 34
“1 (a– ) [10] , the lesions were irregular in shape and blurred in boundry.”
Shabani, S., Homayounfar, M., Vardhanabhuti, V., Mahani, M.-A. N., & Koohi‐Moghadam, M. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.106033 · Cited by 34
“To train, validate, and test our pipeline, we use four different public datasets, namely MosmedData [32], Coro-naCases [33,34], UESTC [35,36], and Radiopeadia [37].”
MethodFu, H., Wang, G., Lei, W., Xu, W., Zhao, Q., Zhang, S., Li, K., & Zhang, S. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3181462 · Cited by 33
“2) Implementation Details: The proposed framework was implemented in PyMIC3 [41] using PyTorch on one NVIDIA GTX 2080 Ti GPU with 12 GB memory.”
MethodSelf-citationChen, H., Jiang, Y., Ko, H., & Loew, M. H. (2022) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2022.104250 · Cited by 33
Cov-TransNet: Dual branch fusion network with transformer for COVID-19 infection segmentation
Peng, Y., Zhang, T., & Guo, Y. (2022) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2022.104366 · Cited by 33
Semi-Supervised CT Lesion Segmentation Using Uncertainty-Based Data Pairing and SwapMix
Qiao, P., Li, H., Song, G., Han, H., Gao, Z., Tian, Y., Liang, Y., Li, X., Zhou, S. K., & Chen, J. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3232572 · Cited by 33
“Wang et al. [56] proposed a noise-robust COPLE-Net to delineate COVID-19 lesions from CT images with noisy labels.”
BackgroundZhong, L., Wang, G., Liao, X., & Zhang, S. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3269798 · Cited by 34
Self-citationLiu, C., Albrecht, C. M., Wang, Y., Li, Q., & Zhu, X. X. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3373908 · Cited by 34
“In the former case from classification tasks, a set of regularization techniques such as consistency regularization [11] or entropy minimization [12] is used to constrain the optimization space.”
BackgroundMethodKarthik, R., Menaka, R., Hariharan, M., & Kathiresan, G. S. (2021) · IRBM · doi:10.1016/j.irbm.2021.07.002 · Cited by 32
“[195] CT Adaptive self-ensembling CNN 130 CT scans 3 Fan et al.”
Multi-Scale Interactive Network With Artery/Vein Discriminator for Retinal Vessel Classification
Hu, J., Wang, H., Wu, G., Cao, Z., Mou, L., Zhao, Y., & Zhang, J. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3165867 · Cited by 32
“In this work, through experimental verification, the NRDice Loss [34]”
MethodPunitha, S. C., Stephan, T., Kannan, R., Mahmud, M., Kaiser, M. S., & Belhouari, S. B. (2023) · IEEE Access · doi:10.1109/access.2023.3236812 · Cited by 32
“Wang et al. [48], proposed an automated system for extracting pneumonia lesions from CT images.”
MethodZhang, C., Bahrami, M., Mishra, D. K., Yuen, M. M. F., Yu, Y., & Zhang, J. (2025) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2025.110909 · Cited by 31
RETRACTED ARTICLE: Modeling the progression of COVID-19 deaths using Kalman Filter and AutoML
Han, T., Gois, F. N. B., Oliveira, R., Prates, L. R., & Porto, M. M. D. A. (2021) · Soft Computing - A Fusion of Foundations, Methodologies and Applications · doi:10.1007/s00500-020-05503-5 · Cited by 18
RetractedH3DNN: 3D Deep Learning Based Detection of COVID-19 Virus using Lungs Computed Tomography
Khan, A. A., Shafiq, S., Kumar, R., Kumar, J., & Haq, A. U. (2020) · International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP) · doi:10.1109/iccwamtip51612.2020.9317357 · Cited by 30
“Over the past few months, many deep learning based screening approaches have been proposed to detect infected lesions from 2D CT imaging [7-9].”
BackgroundMPS-Net: Multi-Point Supervised Network for CT Image Segmentation of COVID-19
Pei, H.-Y., Yang, D., Liu, G.-R., & Lu, T. (2021) · IEEE Access · doi:10.1109/access.2021.3067047 · Cited by 30
“[25] proposed a COPLE-Net automatic segmentation network and a new loss function (NR-Dice).”
MEDIAR: Harmony of Data-Centric and Model-Centric for Multi-Modality MicroscopyPreprint
Lee, G., Kim, S., Kim, J., & Yun, S.-Y. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2212.03465 · Cited by 30
Data efficient deep learning for medical image analysis: A surveyPreprint
Kumari, S., & Singh, P. P. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2310.06557 · Cited by 30
“In addition, some works choose to adapt loss functions specifically for tasks in medical imaging [284, 285, 286].”
BackgroundIntelligent Health Monitoring System for Detection of Symptomatic/Asymptomatic COVID-19 Patient
Nandy, S., & Adhikari, M. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2021.3096425 · Cited by 29
An optimized two-stage cascaded deep neural network for adrenal segmentation on CT images
Luo, G., Yang, Q., Chen, T., Zheng, T., Xie, W., & Sun, H. (2021) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2021.104749 · Cited by 29
Structural Attention Graph Neural Network for Diagnosis and Prediction of COVID-19 Severity
Liu, Y., Li, H., Luo, T., Zhang, C., Xiao, Z., Wei, Y., Gao, Y., Shi, F., Shan, F., & Shen, D. G. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3226575 · Cited by 29
“The bulk of studies focus on joint segmentation and localization of COVID-19 lesions [34], [35], [36], [37].”
BackgroundRobust T-Loss for Medical Image Segmentation
González-Jiménez, Á., Lionetti, S., Gottfrois, P., Gröger, F., Pouly, M., & Navarini, A. A. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43898-1_68 · Cited by 28
Label Noise in Segmentation Networks: Mitigation Must Deal with Bias
Vorontsov, E., & Kadoury, S. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88210-5_25 · Cited by 28
A Survey on Machine Learning in COVID-19 Diagnosis
Guo, X. L., Zhang, Y., Lu, S., & Lu, Z. (2021) · Computer Modeling in Engineering & Sciences · doi:10.32604/cmes.2021.017679 · Cited by 27
Ding, W., Abdel‐Basset, M., Hawash, H., & Elkomy, O. M. (2021) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2021.3123173 · Cited by 27
“Moreover, COPLE-Net [16] and Inf-Net [14] achieved comparable performance with DSC of 0.849 and 0.851, respectively.”
BackgroundInfluentialRegion-based online selective examination for weakly supervised semantic segmentation
Chen, Q., Chen, Y., Huang, Y., Xie, X., & Yang, L. (2024) · Information Fusion · doi:10.1016/j.inffus.2024.102311 · Cited by 27
Deep Learning-Based Building Footprint Extraction With Missing Annotations
Kang, J. S., Fernández-Beltrán, R., Sun, X., Ni, J., & Plaza, A. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3072589 · Cited by 27
“Although both NR-Dice and WDice are designed for the tackling segmentation problem with unbalanced label distribution, one plausible reason that explains why our joint loss outperforms those is the fact that it considers the penalization of boundary misalignments.”
MethodSemantic segmentation of COVID-19 lesions with a multiscale dilated convolutional network
Zhang, J., Ding, X., Hu, D., & Jiang, Y. (2022) · Scientific Reports · doi:10.1038/s41598-022-05527-x · Cited by 26
“Wang et al. 16 proposed a novel framework introducing a noise-robust Dice loss function to learn from noisy labels to segment the infected regions from CT images.”
BackgroundDiwakar, M., Singh, P., Swarup, C., Bajal, E., Jindal, M., Ravi, V., Singh, K. U., & Singh, T. (2022) · Diagnostics · doi:10.3390/diagnostics12112766 · Cited by 26
“Primarily, it divides the noisy image into multiple sub-bands oriented to NSST, later using a twin support vector for feature detection and denoising [42].”
Hu, H., Xu, W., Jiang, T., Cheng, Y., Tao, X., Liu, W., Jian, M., Li, K., & Wang, G. (2022) · Clinical Chemistry · doi:10.1093/clinchem/hvac190 · Cited by 26
Self-citationMWG-Net: Multiscale Wavelet Guidance Network for COVID-19 Lung Infection Segmentation From CT Images
Hu, K., Tan, H. Y., Zhang, Y., Huang, W., & Gao, X. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3265100 · Cited by 26
“In order to improve the robustness of the network, Wang et al. [33] proposed a noise-robust loss function and a novel COVID-19 pneumonia lesion network (COPLE-Net), which combines two different pooling operations and bridge layers.”
BackgroundA weakly supervised inpainting-based learning method for lung CT image segmentation
Lu, F., Zhang, Z., Liu, T., Tang, C., Bai, H., Zhai, G., Chen, J., & Wu, X. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.109861 · Cited by 26
An evolutionary Chameleon Swarm Algorithm based network for 3D medical image segmentation
Rajesh, C., Sadam, R., & Kumar, S. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122509 · Cited by 26
Bose, S., Chowdhury, R. S., Pal, D., Bose, S., Banerjee, B., & Chaudhuri, S. (2023) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2023.11.002 · Cited by 26
“Noise-robust dice loss [49] alleviates the class imbalance”
BackgroundFu, J., Wang, G., Lu, T., Yue, Q., Vercauteren, T. K., Ourselin, S., & Zhang, S. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.111204 · Cited by 26
Self-citationLasker, A., Obaidullah, S. M., Chakraborty, C., & Roy, K. (2022) · SN Computer Science · doi:10.1007/s42979-022-01464-8 · Cited by 26
Guest Editorial: Special Issue on Imaging-Based Diagnosis of COVID-19
Shen, D. G., Gao, Y., Muñoz‐Barrutia, A., DeBuc, D. C., & Percannella, G. (2020) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2020.3008025 · Cited by 26
Enhancement and segmentation of medical images through pythagorean fuzzy sets-An innovative approach
Premalatha, R., & Dhanalakshmi, P. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-07043-5 · Cited by 25
“The ability to recognize contaminated tissue quickly and precisely is critical for optimal patient treatment and survival [3–6].”
BackgroundCoupled Confusion Correction: Learning from Crowds with Sparse Annotations
Zhang, H., Li, S., Zeng, D., Yan, C. C., & Ge, S. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i15.29613 · Cited by 25
TRACER: Extreme Attention Guided Salient Object Tracing NetworkPreprint
Lee, M. S., Shin, W., & Han, S. (2021) · AAAI Conference on Artificial Intelligence · doi:10.48550/arxiv.2112.07380 · Cited by 118
“Moreover, to further improve the network for equivariance learning and to reduce the divergence discrepancy, we measure the L1 distance, which enables the network to learn robustly against noisy labels (Ghosh, Kumar, and Sastry 2017; Wang et al. 2020a).”
BackgroundMethodSeverity Assessment of COVID-19 Based on Feature Extraction and V-Descriptors
Ye, B., Yuan, X., Cai, Z., & Lan, T. (2021) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2021.3056386 · Cited by 24
Robust Point Cloud Segmentation With Noisy Annotations
Ye, S., Chen, D., Han, S., & Liao, J. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3225323 · Cited by 25
Bizopoulos, P. A., Vretos, N., & Daras, P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2009.06412 · Cited by 24
Diagnosis of COVID-19 Using Machine Learning and Deep Learning: A Review
Mondal, M. R. H. (2021) · Current Medical Imaging Formerly Current Medical Imaging Reviews · doi:10.2174/18756603mte2dnjccy · Cited by 23
Distribution-Aware Margin Calibration for Semantic Segmentation in Images
Yu, L., Li, Z., Xu, M. G., Gao, Y., Luo, J., & Zhang, J. (2021) · International Journal of Computer Vision · doi:10.1007/s11263-021-01533-0 · Cited by 24
Ding, W., Abdel‐Basset, M., Hawash, H., Pratama, M., & Pedrycz, W. (2023) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2023.3245103 · Cited by 23
“4) UESTC-COVID-19 Dataset consists of 50 CT volumes annotated with multiple domain experts [29].”
BackgroundMethodDu, J., Zhang, X., Liu, P., & Wang, T. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2023.3278741 · Cited by 23
“MT has been applied in medical image analysis, such as skinlesionandliversegmentation[30]orCOVID-19pneumonia lesionsegmentationfromCTscanswithnoisylabels[31].”
MethodBougourzi, F., Dornaika, F., Ahmed, A. T., & Hoang, V. T. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-78128-5_16 · Cited by 23
Mohammadi, A., Wang, Y., Enshaei, N., Afshar, P., Naderkhani, F., Oikonomou, A., Rafiee, J., Oliveira, H. C. R., Yanushkevich, S. N., & Plataniotis, K. N. (2021) · IEEE Signal Processing Magazine · doi:10.1109/msp.2021.3090674 · Cited by 23
Kufel, J., Bargieł-Łączek, K., Koźlik, M., Czogalik, Ł., Dudek, P., Jaworski, A., Cebula, M., & Gruszczyńska, K. (2022) · International Journal of Medical Sciences · doi:10.7150/ijms.76515 · Cited by 23
“Currently, the gold standard for diagnosing these patients is the reverse transcription–polymerase chain reaction (RT-PCR) test [1,2].”
BackgroundHussain, M. A., Mirikharaji, Z., Momeny, M., Marhamati, M., Neshat, A. A., Garbi, R., & Hamarneh, G. (2022) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2022.102127 · Cited by 22
“…al. (2021), Abdel-Basset et al. (2020), Yang et al. (2021), Taghanaki et al. (2019a), Souly et al. (2017), Lee et al. (2019), Hong et al. (2015) and Wang et al. (2020), however, native 3D segmentation would be superior to the reconstruction of 3D segmentation from stacked slice-based 2D…”
Casting defect region segmentation method based on dual-channel encoding–fusion decoding network
Jiang, H., Zhang, X., Tao, C., Ai, S., Wang, Y., He, J., He, Y., & Yang, D. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.123254 · Cited by 22
Ding, W., Abdel‐Basset, M., & Hawash, H. (2021) · Information Sciences · doi:10.1016/j.ins.2021.07.059 · Cited by 22
“To tackle these limitations, annotation-efficient DL for medical image segmentation has attracted growing research to relax the requirement of a large-scale, pixel-level annotated CT dataset for training, inspiring training techniques with partial or no supervision, such as unsupervised domain adaptation (UDA) [9], we…”
InfluentialRahaman, M., Hillas, M. M., Tuba, J., Ruma, J. F., Ahmed, N., & Rahman, R. M. (2021) · Cybernetics & Systems · doi:10.1080/01969722.2021.1989171 · Cited by 22
Towards Unbiased Covid-19 Lesion Localisation And Segmentation Via Weakly Supervised Learning
Yang, Y., Chen, J., Wang, R., Ma, T., Wang, L., Chen, J., Zheng, W., & Zhang, T. (2021) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi48211.2021.9433806 · Cited by 21
“COPLE-Net was designed to enhance the robustness of the detection, using the labels polluted by the noise data from non-experts [9].”
MethodFDA: Feature Decomposition and Aggregation for Robust Airway Segmentation
Zhang, M., Xin, Y., Zhang, H., Zheng, H., Yu, W., Pan, H., Cai, X., & Gu, Y. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87722-4_3 · Cited by 21
Bougourzi, F., Dornaika, F., Nakib, A., & Ahmed, A. T. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10717-2 · Cited by 21
“First, the available labelled data are very limited due to the significant time and effort required for the labelling process, with physicians and radiologists being overwhelmed due to the pandemic (Wang et al. 2020; Pezzano et al. 2021; Yao et al. 2021).”
3D medical image segmentation based on semi-supervised learning using deep co-training
Yang, J., Li, H., Wang, H., & Man, H. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.111641 · Cited by 21
Learning trustworthy model from noisy labels based on rough set for surface defect detection
Niu, T., Wang, Z., Li, W., Li, K., Li, K., Xu, G., & Li, Y. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.112138 · Cited by 21
Self-citationHuang, H., Cai, M., Lin, L., Zheng, J., Mao, X., Qian, X., Peng, Z., Zhou, J., Iwamoto, Y., Han, X., et al. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9413957 · Cited by 21
“The classic UNet [6] and UNet++ [7] were widely performed as segmentation architectures for COVID-19.”
InfluentialQayyum, A., Mazhar, M., Razzak, I., & Bouadjenek, M. R. (2021) · Neural Computing and Applications · doi:10.1007/s00521-021-06636-w · Cited by 21
“Recently, deep learning has been extensively use for medical image analysis, and several deep learning methods have presented to analyze the Chest images (Ultrasound, CT and MRI) for the detection of patients infected with COVID-19 [4, 8, 29].”
BackgroundMethodShah, A., & Shah, M. (2022) · Chronic Diseases and Translational Medicine · doi:10.1002/cdt3.17 · Cited by 21
“Wang et al.(58) COPLE‐Net (Modified U‐Net + ASE) Fully supervised 558 Scans 0.”
BackgroundInfluentialSTCNet: Alternating CNN and improved transformer network for COVID-19 CT image segmentation
Geng, P., Tan, Z., Wang, Y., Jia, W., Zhang, Y., & Yan, H. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106205 · Cited by 20
Yang, Y., Hu, L., Zu, C., Zhang, J., Hou, Y., Chen, Y., Zhou, J., Zhou, L., & Wang, Y. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110741 · Cited by 20
Dang, Y., Ma, W., Luo, X., & Wang, H. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103583 · Cited by 20
Zhao, W., Lou, M., Qi, Y., Wang, Y., Xu, C., Deng, X., & Ma, Y. (2021) · Applied Intelligence · doi:10.1007/s10489-021-02297-3 · Cited by 20
Zhang, M., Zhang, H., Yang, G., & Gu, Y. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16431-6_48 · Cited by 20
Asad, M., Fidon, L., & Vercauteren, T. K. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2201.04584 · Cited by 20
“Deep learning-based automatic lung lesion segmentation methods may ease burden on clinicians, however, these methods require large amounts of manually labelled data (Wang et al., 2020; Gonzalez et al., 2021; Tilborghs et al., 2020; Chassagnon et al., 2020).”
BackgroundMethodInfluentialHDConv: Heterogeneous kernel-based dilated convolutions
Hu, H., Yu, C.-H., Zhou, Q., Guan, Q., & Feng, H. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106568 · Cited by 19
Liver segmentation network based on detail enhancement and multi-scale feature fusion
Tinglan, L., Qin, J., Guihe, Q., Shi, W., & Zhang, W. (2025) · Scientific Reports · doi:10.1038/s41598-024-78917-y · Cited by 19
“For instance, Wang et al. constructed a multi-scale context-aware module in the encoder of MSR-Net 32 , learning contextual information of target regions from multiple spatial and channel scales to enhance the semantic representation capability of the network.”
InfluentialMeddeb, A., Kossen, T., Bressem, K. K., Hamm, B., & Nagel, S. N. (2021) · Tomography · doi:10.3390/tomography7040078 · Cited by 19
Akbar, S. B., Thanupillai, K., & Sundararaj, S. (2022) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.6958 · Cited by 18
Zhang, T., & Wang, X. (2023) · IEEE Transactions on Fuzzy Systems · doi:10.1109/tfuzz.2023.3298904 · Cited by 18
“AsshowninTableIII,thereare fourpopularFCNs,i.e.,theU-Net++[6],attentionU-Net(ATT-UNet) [10], spatial and channel squeeze & excitation network (ScSE-Net) [9], and COPLE-Net [11].”
BackgroundSenseCare: a research platform for medical image informatics and interactive 3D visualization
Wang, G., Duan, Q., Tian, S., & Zhang, S. (2024) · Frontiers in Radiology · doi:10.3389/fradi.2024.1460889 · Cited by 18
Self-citationPezzano, G., Díaz, O., Ribas, V. J., & Radeva, P. I. (2021) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2021.104689 · Cited by 18
AGMR-Net: Attention-guided multiscale recovery framework for stroke segmentation
Du, X., Ma, K., & Song, Y. (2022) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2022.102120 · Cited by 18
“Unet++ and CLCI-Net, although they modify the skip connection in the encoding stage to achieve better recovery of the target The effect of boundary, Deeplab V3 also uses multi-scale infor- mation, but it is still insufficient in the boundary segmentation effect.”
BackgroundMethodInfluentialYang, S., Wang, G., Sun, H., Luo, X., Sun, P., Li, K., Wang, Q., & Zhang, S. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3172978 · Cited by 18
“The first section in Table VI shows the quantitative results of these methods with the clean label percentage of 15 % , where our framework achieves significant improvements compared to the second best method (i.e., NRF [27]).”
BackgroundMethodInfluentialSelf-citationCOVID-19 and Associated Lung Disease Classification Using Deep Learning
Bhosale, Y. H., Singh, P., & Patnaik, K. S. (2022) · Lecture notes in networks and systems · doi:10.1007/978-981-19-3679-1_22 · Cited by 18
MID-UNet: Multi-input directional UNet for COVID-19 lung infection segmentation from CT images
Chi, J., Zhang, S., Han, X., Wang, H., Wu, C., & Yu, X. (2022) · Signal Processing Image Communication · doi:10.1016/j.image.2022.116835 · Cited by 17
Self-citationBoundary Aware U-Net for Medical Image Segmentation
Alahmadi, M. D. (2022) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-022-07431-y · Cited by 17
“COPLE-Net [52] and SAR-U-Net [53] take advantage of the SE module for Lung and Liver segmentation, respectively.”
BackgroundThe Role of Data Quality for Reliable AI Performance in Medical Applications
Enshaei, N., & Naderkhani, F. (2024) · IEEE reliability magazine · doi:10.1109/mrl.2024.3430192 · Cited by 17
“To address this challenge, several approaches have been suggested, including extracting pathology labels from large-scale medical images using text mining techniques [9], employing human–machine collaborative strategies [10], and utilizing nonexpert annotators [11].”
BackgroundMethodAlam, M. S., Wang, D., Arzhaeva, Y., Ende, J. A., Kao, J., Silverstone, L., Yates, D. H., Salvado, O., & Sowmya, A. (2024) · Scientific Reports · doi:10.1038/s41598-024-79494-w · Cited by 17
Ding, J., Chang, J., Han, R., & Yang, L. (2025) · International Journal of Biomedical Imaging · doi:10.1155/ijbi/9175473 · Cited by 17
“Bertasius G et al. [16] introduced the DeepEdge model, employing the Canny operator for initial edge information extraction, followed by semantic fusion with corresponding network features of varying sizes to enhance edge pixel classification accuracy.”
Fang, Z., Bai, J., Guo, X., Wang, X., Gao, F., Yang, H., Kong, B., Hou, Y. M., Cao, K., Song, Q., et al. (2022) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2022.3147440 · Cited by 17
“While [7], [16], [18], and [22] all try to alleviate the need for expensive annotation of COVID lesion segmentation, they achieve this goal by different approaches.”
BackgroundMethodInfluentialWang, Z., & Voiculescu, I. D. (2023) · Applied Sciences · doi:10.3390/app13137966 · Cited by 17
Brain Tumor Detection and Segmentation using CNN and Optimised Edge Features
Kanagamalliga, S., & Rajappriadharshine, S. (2024) · International Conference on Communications and Signal Processing (ICCSP) · doi:10.1109/iccsp60870.2024.10544304 · Cited by 16
“Unlike linear filters, bilateral filtering would incorporates both spatial proximity and intensity variations between pixels, ensuring that edges and intricate features remain intact throughout the noise reduction process[5][6].”
BackgroundLearning robust medical image segmentation from multi-source annotations
Wang, Y., Luo, L., Wu, M., Wang, Q., & Chen, H. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103489 · Cited by 16
Zhang, H., Liu, J., Wang, P., Yu, Z., Liu, W., & Chen, H. (2022) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2022.3153793 · Cited by 16
“To improve network robustness to potential noise data, referring to [30], we introduce the Noise-Robust Dice (NRD) loss to train DCSN”
MethodQuery-based black-box attack against medical image segmentation model
Li, S., Huang, G., Xu, X., & Lu, H. (2022) · Future Generation Computer Systems · doi:10.1016/j.future.2022.03.008 · Cited by 16
Levy Flight and Chaos Theory-Based Gravitational Search Algorithm for Image Segmentation
Rather, S. A., & Das, S. R. (2023) · Mathematics · doi:10.3390/math11183913 · Cited by 16
Metadata
- Authors
- Wang, G., Liu, X., Li, C., Xu, Z., Ruan, J., Zhu, H., Meng, T., Li, K., Huang, N., & Zhang, S.
- Affiliations
- University of Electronic Science and Technology of China (CN)
- Sensetime (China) (CN)
- Pingliang People's Hospital (CN)
- Dingxi City People's Hospital (CN)
- Aviation General Hospital (CN)
- Sichuan University (CN)
- Venue
- IEEE Transactions on Medical Imaging
- Published
- 2020-08
- Volume
- 39
- Issue
- 8
- Pages
- 2653-2663
- ISSN
- 0278-0062 (ISSN-L)
- 1558-254X
- PMC
- PMC8544954
- OpenAlex
- W3033272814
- Cited by
- 469
- References
- 57
- Citations by year
- 2026: 23 · 2025: 36 · 2024: 68 · 2023: 88 · 2022: 112 · 2021: 122 · 2020: 20
- FWCI
- 41.62
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Licence
- Locations
- IEEE Transactions on Medical Imaging (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- Funders
- National Natural Science Foundation of China (81771921, 61901084)
- Topics
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › COVID-19 diagnosis using AI
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- COVID-19 pneumonia, lesion segmentation, CT images, label noise, self-ensembling, exponential moving average, medical image segmentation
- MeSH terms
- Betacoronavirus, Deep Learning, COVID-19, SARS-CoV-2, Algorithms, Humans, Lung, Pneumonia, Viral, Tomography, X-Ray Computed, Coronavirus Infections, Pandemics
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.