Quadratic Autoencoder (Q-AE) for Low-Dose CT Denoising
Open accessThe DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index.
Abstract
Inspired by complexity and diversity of biological neurons, our group proposed quadratic neurons by replacing the inner product in current artificial neurons with a quadratic operation on input data, thereby enhancing the capability of an individual neuron. Along this direction, we are motivated to evaluate the power of quadratic neurons in popular network architectures, simulating human-like learning in the form of "quadratic-neuron-based deep learning". Our prior theoretical studies have shown important merits of quadratic neurons and networks in representation, efficiency, and interpretability. In this paper, we use quadratic neurons to construct an encoder-decoder structure, referred as the quadratic autoencoder, and apply it to low-dose CT denoising. The experimental results on the Mayo low-dose CT dataset demonstrate the utility and robustness of quadratic autoencoder in terms of image denoising and model efficiency. To our best knowledge, this is the first time that the deep learning approach is implemented with a new type of neurons and demonstrates a significant potential in the medical imaging field.
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.
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 14×
“For the hyperparameters in WGAN-VGG, we used the configuration recommended in [42].”
BackgroundMethodInfluentialVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 1×
“Numerous excellent models, such as autoencoder [7], VGG [8], Res-Net [9], or GAN [10], were developed, and achieved the state-of-the-art performance, even outperforming human experts in a number of significant tasks, which are frequently reported by media.”
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“eep learning [1-2] has achieved great successes in many important fields including medical imaging over the past several years [3-6].”
BackgroundAdam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 3×
“The Adam optimization was applied for training in the TensorFlow.”
BackgroundLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Densely Connected Convolutional Networks
Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.243 · Cited by 48,566
Pan, S. J., & Yang, Q. (2010) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2009.191 · Cited by 23,387 · cited 1×
“Transfer learning [49] is a type of machine learning techniques with which a model developed for a task can be”
BackgroundReducing the dimensionality of data with neural networks
Hinton, G. E., & Salakhutdinov, R. R. (2006) · Science · doi:10.1126/science.1127647 · Cited by 20,928
Auto-Encoding Variational Bayes
Kingma, D. P., & Welling, M. (2013) · arXiv · doi:10.48550/arxiv.1312.6114 · Cited by 20,067
Labaca-Castro, R. (2023) · Machine Learning under Malware Attack · doi:10.1007/978-3-658-40442-0_9 · Cited by 19,272
GAN(Generative Adversarial Nets)
(2017) · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics · doi:10.3156/jsoft.29.5_177_2 · Cited by 19,270
$rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation
Aharon, M., Elad, M., & Bruckstein, A. (2006) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2006.881199 · Cited by 9,451 · cited 1×
“Albeit clear improvements in certain areas, traditional image post-processing techniques such as non-local means [36], kSVD [37], block-matching 3D [38] are often subject to oversmoothness and structural distortion.”
BackgroundComputed Tomography — An Increasing Source of Radiation Exposure
Brenner, D. J., & Hall, E. J. (2007) · New England Journal of Medicine · doi:10.1056/nejmra072149 · Cited by 8,748 · cited 1×
“Although CT provides critical clinical information, there are potential risks (possible cancer diseases and genetic damages) induced by X-ray radiation [28].”
Background(2010) · Journal of Machine Learning Research · doi:10.5555/1756006.1953039 · Cited by 4,051 · cited 3×
“Among the autoencoders, denoising autoencoders [16] were developed based on the idea that a good and robust representation can be adaptively learned from corrupted data themselves.”
BackgroundMethodPerceptrons: An Introduction to Computational Geometry
Minsky, M., & Papert, S. A. (1969) · Cited by 3,751
Deep Convolutional Neural Network for Inverse Problems in Imaging
Jin, K. H., McCann, M. T., Froustey, E., & Unser, M. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2713099 · Cited by 2,529 · cited 1×
“eep learning [1-2] has achieved great successes in many important fields including medical imaging over the past several years [3-6].”
BackgroundSidky, E. Y., & Pan, X. (2008) · Physics in Medicine and Biology · doi:10.1088/0031-9155/53/17/021 · Cited by 1,995
Low-Dose CT With a Residual Encoder-Decoder Convolutional Neural Network
Chen, H., Zhang, Y., Kalra, M. K., Lin, F., Chen, Y., Liao, P., Zhou, J., & Wang, G. (2017) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2017.2715284 · Cited by 1,866 · cited 4×
“In the experiments, we selected the algorithms CNN10 [39], WGAN-VGG [42], RED-CNN [40], CPCE-2D [43] and BM3D [38] as the baselines in comparison with our quadratic autoencoder.”
BackgroundMethodInfluentialSelf-citationYang, Q., Yan, P., Zhang, Y., Yu, H., Shi, Y., Mou, X., Kalra, M. K., Zhang, Y., Sun, L., & Wang, G. (2018) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2018.2827462 · Cited by 1,630 · cited 9×
“For the hyperparameters in WGAN-VGG, we used the configuration recommended in [42].”
MethodInfluentialSelf-citationContractive Auto-Encoders: Explicit Invariance During Feature Extraction
Rifai, S., Vincent, P., Muller, X., Glorot, X., & Bengio, Y. (2011) · International Conference on Machine Learning · Cited by 1,603
Makhzani, A., Shlens, J., Jaitly, N., Goodfellow, I., & Frey, B. (2015) · arXiv · doi:10.48550/arxiv.1511.05644 · Cited by 1,579
Generative Adversarial Networks for Noise Reduction in Low-Dose CT
Wolterink, J. M., Leiner, T., Viergever, M. A., & Isgum, I. (2017) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2017.2708987 · Cited by 1,037 · cited 2×
“This paper started a fast development of this area, involving the similarity loss, perceptual loss, adversarial mechanism, wavelet, framelet, and transfer learning, as well as various network structures, published as RED-CNN [40], GAN-3D [41], WGAN-VGG [42], CPCE-2D [43], wavelet residual networks [44], SMGAN-3D [45]…”
(2014) · arXiv (Cornell University) · Cited by 942
aLow-dose CT via convolutional neural network
Chen, H., Zhang, Y., Zhang, W., Liao, P., Li, K., Zhou, J., & Wang, G. (2017) · Biomedical Optics Express · doi:10.1364/boe.8.000679 · Cited by 755 · cited 10×
“RED-CNN and Q-AE perform much better than CNN10, CPCE-2D and WGAN-VGG.”
MethodInfluentialSelf-citationA deep convolutional neural network using directional wavelets for low‐dose X‐ray CT reconstruction
Kang, E., Min, J., & Ye, J. C. (2017) · Medical Physics · doi:10.1002/mp.12344 · Cited by 709 · cited 1×
“published as RED-CNN [40], GAN-3D [41], WGAN-VGG [42], CPCE-2D [43], wavelet residual networks [44], SMGAN-3D [45] and so on.”
Low-Dose X-ray CT Reconstruction via Dictionary Learning
Xu, Q., Yu, H., Mou, X., Zhang, L., Hsieh, J., & Wang, G. (2012) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2012.2195669 · Cited by 690
Self-citationPenalized weighted least-squares image reconstruction for dual energy X-ray transmission tomography
Sukovic, P., & Clinthorne, N. H. (2000) · IEEE Transactions on Medical Imaging · doi:10.1109/42.896783 · Cited by 552
Wang, J., Li, T., Lu, H., & Liang, Z. (2006) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2006.882141 · Cited by 533 · cited 1×
“To cope with such image degradation, a plethora of algorithms have been developed, which can be classified into the three categories: (i) sinogram filtering [29-31] (ii) iterative reconstruction [32-35], and (iii) image post-processing.”
MethodLEARN: Learned Experts’ Assessment-Based Reconstruction Network for Sparse-Data CT
Chen, H., Zhang, Y., Chen, Y., Zhang, J., Zhang, W., Sun, H., Lv, Y., Liao, P., Zhou, J., & Wang, G. (2018) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2018.2805692 · Cited by 433 · cited 1×
“eep learning [1-2] has achieved great successes in many important fields including medical imaging over the past several years [3-6].”
BackgroundSelf-citationShan, H., Padole, A., Homayounieh, F., Kruger, U., Khera, R. D., Nitiwarangkul, C., Kalra, M. K., & Wang, G. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0057-9 · Cited by 417 · cited 1×
“eep learning [1-2] has achieved great successes in many important fields including medical imaging over the past several years [3-6].”
BackgroundSelf-citationShan, H., Zhang, Y., Yang, Q., Kruger, U., Kalra, M. K., Sun, L., Cong, W., & Wang, G. (2018) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2018.2832217 · Cited by 411 · cited 2×
“In the experiments, we selected the algorithms CNN10 [39], WGAN-VGG [42], RED-CNN [40], CPCE-2D [43] and BM3D [38] as the baselines in comparison with our quadratic autoencoder.”
MethodSelf-citationProjection space denoising with bilateral filtering and CT noise modeling for dose reduction in CT
Manduca, A., Yu, L., Trzasko, J. D., Khaylova, N., Kofler, J. M., McCollough, C. M., & Fletcher, J. G. (2009) · Medical Physics · doi:10.1118/1.3232004 · Cited by 360
Li, Z., Yu, L., Trzasko, J. D., Lake, D. S., Blezek, D. J., Fletcher, J. G., McCollough, C. H., & Manduca, A. (2013) · Medical Physics · doi:10.1118/1.4851635 · Cited by 299 · cited 1×
“Albeit clear improvements in certain areas, traditional image postprocessing techniques such as non-local means [36], k-SVD [37], block-matching 3D [38] are often subject to oversmoothness and structural distortion.”
BackgroundPerceptrons: An Introduction to Computational Geometry
Minsky, M., & Papert, S. A. (2017) · The MIT Press eBooks · doi:10.7551/mitpress/11301.001.0001 · Cited by 271
Structurally-Sensitive Multi-Scale Deep Neural Network for Low-Dose CT Denoising
You, C., Cong, W., Wang, G., Yang, Q., Shan, H., Gjesteby, L., Li, G., Ju, S., Zhang, Z., Zhao, Z., et al. (2018) · IEEE Access · doi:10.1109/access.2018.2858196 · Cited by 259 · cited 2×
“This paper started a fast development of this area, involving the similarity loss, perceptual loss, adversarial mechanism, wavelet, framelet, and transfer learning, as well as various network structures, published as RED-CNN [40], GAN-3D [41], WGAN-VGG [42], CPCE-2D [43], wavelet residual networks [44], SMGAN-3D [45]…”
Self-citationOn the Computational Efficiency of Training Neural Networks
Livni, R., Shalev‐Shwartz, S., & Shamir, O. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1410.1141 · Cited by 245 · cited 1×
“[23] altered the activation function of common artificial neurons to a quadratic form: σ(z) = z(2), but such a nonlinearity is only of limited utility because at the cellular level the decision boundary is still linear.”
BackgroundUsing neuroscience to develop artificial intelligence
Ullman, S. (2019) · Science · doi:10.1126/science.aau6595 · Cited by 231 · cited 1×
“Nevertheless, typical neural networks only employ simplified homogeneous neurons [11].”
BackgroundBlock matching 3D random noise filtering for absorption optical projection tomography
Fumene Feruglio, P., Vinegoni, C., Gros, J., Sbarbati, A., & Weissleder, R. (2010) · Physics in Medicine and Biology · doi:10.1088/0031-9155/55/18/009 · Cited by 175 · cited 4×
“In the experiments, we selected the algorithms CNN10 [39], WGAN-VGG [42], RED-CNN [40], CPCE-2D [43] and BM3D [38] as the baselines in comparison with our quadratic autoencoder.”
BackgroundMethodInfluentialRay Contribution Masks for Structure Adaptive Sinogram Filtering
Balda, M., Hornegger, J., & Heismann, B. J. (2012) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2012.2187213 · Cited by 156 · cited 1×
“To cope with such image degradation, a plethora of algorithms have been developed, which can be classified into the three categories: (i) sinogram filtering [29-31] (ii) iterative reconstruction [32-35], and (iii) image post-processing.”
MethodCai, J., Jia, X., Gao, H., Jiang, S. B., Shen, Z., & Zhao, H. (2014) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2014.2319055 · Cited by 154 · cited 1×
“To cope with such image degradation, a plethora of algorithms have been developed, which can be classified into the three categories: (i) sinogram filtering [29-31] (ii) iterative reconstruction [32-35], and (iii) image post-processing.”
MethodNon-linear Convolution Filters for CNN-Based Learning
Zoumpourlis, G., Doumanoglou, A., Vretos, N., & Daras, P. (2017) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.510 · Cited by 124 · cited 1×
“replaced micro-networks with polynomial kernels [26].”
Universal approximation with quadratic deep networks
Fan, F., Xiong, J., & Wang, G. (2020) · Neural Networks · doi:10.1016/j.neunet.2020.01.007 · Cited by 111
Self-citationDense Associative Memory Is Robust to Adversarial Inputs
Krotov, D., & Hopfield, J. (2018) · Neural Computation · doi:10.1162/neco_a_01143 · Cited by 106 · cited 1×
“In the literature, there are two kinds of quadratic activations: the standard one, y = αx [23] and rectified quadratic activation [52]:”
BackgroundA new type of neurons for machine learning
Fan, F., Cong, W., & Wang, G. (2017) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2920 · Cited by 81 · cited 1×
“In reference to the diversity and complexity of biological neurons and human-like learning, our group recently proposed quadratic neurons [12-15] by replacing the inner product in popular artificial neurons with a quadratic operation on input data, thereby enhancing the processing capability of an individual neuron.”
BackgroundSelf-citationKATE: K-Competitive Autoencoder for Text
Chen, Y., & Zaki, M. J. (2017) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3097983.3098017 · Cited by 79 · cited 2×
“K-competitive autoencoders [21] force neurons in the middle layer to compete for the right of response to the input data so that each neuron is specialized.”
BackgroundMajumdar, A. (2018) · Neural Networks · doi:10.1016/j.neunet.2018.07.016 · Cited by 53 · cited 2×
“Graph autoencoders [22] use graph regularization so that the local consistency of a low dimensional manifold can be integrated into representation learning.”
BackgroundSORT: Second-Order Response Transform for Visual Recognition
Wang, Y., Xie, L., Liu, C., Qiao, S., Zhang, Y., Zhang, W., Tian, Q., & Yuille, A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.152 · Cited by 51
Fuzzy logic interpretation of quadratic networks
Fan, F., & Wang, G. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.09.001 · Cited by 29 · cited 1×
“In reference to the diversity and complexity of biological neurons and human-like learning, our group recently proposed quadratic neurons [12]–[15] by replacing the inner product in popular artificial neurons with a quadratic operation on input data, thereby enhancing the processing capability of an…”
BackgroundSelf-citationGeneralized backpropagation algorithm for training second‐order neural networks
Fan, F., Cong, W., & Wang, G. (2018) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2956 · Cited by 29
Self-citationNeurons With Paraboloid Decision Boundaries for Improved Neural Network Classification Performance
Tsapanos, N., Tefas, A., Nikolaidis, N., & Pitas, I. (2018) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2839655 · Cited by 22 · cited 1×
“[24] suggested a parabolic neuron, which is a special case of a quadratic neuron.”
Background
Cited by
Content-Noise Complementary Learning for Medical Image Denoising
Geng, M., Meng, X., Yu, J., Zhu, L., Jin, L., Jiang, Z., Qiu, B., Li, H., Kong, H., Yuan, J., et al. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2021.3113365 · Cited by 183
Self-citation
Metadata
- Authors
- Fan, F., Shan, H., Kalra, M. K., Singh, R., Qian, G., Getzin, M., Teng, Y., Hahn, J., & Wang, G.
- Venue
- IEEE Transactions on Medical Imaging
- Published
- 2020-06
- Volume
- 39
- Issue
- 6
- Pages
- 2035-2050
- ISSN
- 0278-0062 (ISSN-L)
- 1558-254X
- OpenAlex
- W2964296115
- Cited by
- 159
- References
- 50
- FWCI
- 4.02
- Type
- journal article
- Language
- en
- Open access
- green
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.