Unsupervised Training of Denoisers for Low-Dose CT Reconstruction Without Full-Dose Ground Truth

Recently, deep neural network (DNN) based methods for low-dose CT have been investigated to achieve excellent performance in both image quality and computational speed. However, almost all methods using DNNs for low-dose CT require clean ground truth data with full radiation dose to train the DNNs. In this work, we attempt to train DNNs for low-dose CT reconstructions with reduced tube current by investigating unsupervised training of DNNs for denoising sensor measurements or sinograms without full-dose ground truth images. In other words, our proposed methods allow training of DNNs with only noisy low-dose CT measurements. First, the Poisson Unbiased Risk Estimator (PURE) is investigated to train a DNN for denoising CT measurements, and a method is proposed for reconstructing the CT image using filtered back-projection (FBP) and the DNN trained with PURE. Then, the CT forward model-based Weighted Stein's Unbiased Risk Estimator (WSURE) is proposed to train a DNN for denoising CT sinograms and to subsequently reconstruct the CT image using FBP. Our proposed methods achieve excellent performance in both fast computation and reconstructed image quality, which is more comparable to the results of the DNNs trained with full-dose ground truth data than other state-of-the-art denoising methods such as the BM3D, Deep Image Prior, and Deep Decoder.

[1]  Yi Zhang,et al.  Low-Dose CT via Deep CNN with Skip Connection and Network in Network , 2018, Developments in X-Ray Tomography XII.

[2]  Xuanqin Mou,et al.  Low-Dose CT Image Denoising Using a Generative Adversarial Network With Wasserstein Distance and Perceptual Loss , 2017, IEEE Transactions on Medical Imaging.

[3]  Jaejun Yoo,et al.  Deep Convolutional Framelet Denosing for Low-Dose CT via Wavelet Residual Network , 2017, IEEE Transactions on Medical Imaging.

[4]  Lei Zhang,et al.  Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising , 2016, IEEE Transactions on Image Processing.

[5]  Zhaoying Bian,et al.  Optimizing a Parameterized Plug-and-Play ADMM for Iterative Low-Dose CT Reconstruction , 2019, IEEE Transactions on Medical Imaging.

[6]  Se Young Chun,et al.  Training deep learning based image denoisers from undersampled measurements without ground truth and without image prior — Supplementary Material — , 2019 .

[7]  Eero P. Simoncelli,et al.  Learning to be Bayesian without Supervision , 2006, NIPS.

[8]  Jaakko Lehtinen,et al.  Noise2Noise: Learning Image Restoration without Clean Data , 2018, ICML.

[9]  Jong Chul Ye,et al.  Framing U-Net via Deep Convolutional Framelets: Application to Sparse-View CT , 2017, IEEE Transactions on Medical Imaging.

[10]  Jeffrey A. Fessler,et al.  Combining Ordered Subsets and Momentum for Accelerated X-Ray CT Image Reconstruction , 2015, IEEE Transactions on Medical Imaging.

[11]  Michael Unser,et al.  Deep Convolutional Neural Network for Inverse Problems in Imaging , 2016, IEEE Transactions on Image Processing.

[12]  Hongming Shan,et al.  Super-resolution MRI and CT through GAN-CIRCLE , 2019, Developments in X-Ray Tomography XII.

[13]  Yonina C. Eldar.  Generalized SURE for Exponential Families: Applications to Regularization , 2008, IEEE Transactions on Signal Processing.

[14]  Mathias Unberath,et al.  Deep Learning Computed Tomography: Learning Projection-Domain Weights From Image Domain in Limited Angle Problems , 2018, IEEE Transactions on Medical Imaging.

[15]  Jin Liu,et al.  Artifact Removal using Improved GoogLeNet for Sparse-view CT Reconstruction , 2018, Scientific Reports.

[16]  Jeffrey A. Fessler,et al.  PWLS-ULTRA: An Efficient Clustering and Learning-Based Approach for Low-Dose 3D CT Image Reconstruction , 2017, IEEE Transactions on Medical Imaging.

[17]  L. Eon Bottou.  Online Learning and Stochastic Approximations , 1998 .

[18]  Michael Elad,et al.  DeepRED: Deep Image Prior Powered by RED , 2019, ICCV 2019.

[19]  Ke Li,et al.  Learning to Reconstruct Computed Tomography Images Directly From Sinogram Data Under A Variety of Data Acquisition Conditions , 2019, IEEE Transactions on Medical Imaging.

[20]  E. Sidky,et al.  Image reconstruction in circular cone-beam computed tomography by constrained, total-variation minimization , 2008, Physics in medicine and biology.

[21]  Yonina C. Eldar,et al.  On Divergence Approximations for Unsupervised Training of Deep Denoisers Based on Stein’s Unbiased Risk Estimator , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[22]  Huazhong Shu,et al.  Domain Progressive 3D Residual Convolution Network to Improve Low-Dose CT Imaging , 2019, IEEE Transactions on Medical Imaging.

[23]  Andrea Vedaldi,et al.  Deep Image Prior , 2017, International Journal of Computer Vision.

[24]  Guang Li,et al.  CT Super-Resolution GAN Constrained by the Identical, Residual, and Cycle Learning Ensemble (GAN-CIRCLE) , 2018, IEEE Transactions on Medical Imaging.

[25]  Max A. Viergever,et al.  Generative Adversarial Networks for Noise Reduction in Low-Dose CT , 2017, IEEE Transactions on Medical Imaging.

[26]  Baiyu Chen,et al.  Low‐dose CT for the detection and classification of metastatic liver lesions: Results of the 2016 Low Dose CT Grand Challenge , 2017, Medical physics.

[27]  Jian Zhou,et al.  Low-dose CT count-domain denoising via convolutional neural network with filter loss , 2019, Medical Imaging.

[28]  Ge Wang,et al.  Structurally-Sensitive Multi-Scale Deep Neural Network for Low-Dose CT Denoising , 2018, IEEE Access.

[29]  Pascal Vincent,et al.  Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion , 2010, J. Mach. Learn. Res..

[30]  Feng Lin,et al.  Low-Dose CT With a Residual Encoder-Decoder Convolutional Neural Network , 2017, IEEE Transactions on Medical Imaging.

[31]  Loïc Royer,et al.  Noise2Self: Blind Denoising by Self-Supervision , 2019, ICML.

[32]  Thierry Blu,et al.  Monte-Carlo Sure: A Black-Box Optimization of Regularization Parameters for General Denoising Algorithms , 2008, IEEE Transactions on Image Processing.

[33]  Jongha Lee,et al.  Deep-Neural-Network-Based Sinogram Synthesis for Sparse-View CT Image Reconstruction , 2018, IEEE Transactions on Radiation and Plasma Medical Sciences.

[34]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[35]  C. Stein.  Estimation of the Mean of a Multivariate Normal Distribution , 1981 .

[36]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[37]  Hu Chen,et al.  LEARN: Learned Experts’ Assessment-Based Reconstruction Network for Sparse-Data CT , 2017, IEEE Transactions on Medical Imaging.

[38]  Thierry Blu,et al.  The SURE-LET Approach to Image Denoising , 2007, IEEE Transactions on Image Processing.

[39]  Thomas Pock,et al.  Learning a variational network for reconstruction of accelerated MRI data , 2017, Magnetic resonance in medicine.

[40]  Daniel Kolditz,et al.  Iterative reconstruction methods in X-ray CT. , 2012, Physica medica : PM : an international journal devoted to the applications of physics to medicine and biology : official journal of the Italian Association of Biomedical Physics.

[41]  Florian Jug,et al.  Noise2Void - Learning Denoising From Single Noisy Images , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[42]  Jeffrey A. Fessler,et al.  A Splitting-Based Iterative Algorithm for Accelerated Statistical X-Ray CT Reconstruction , 2012, IEEE Transactions on Medical Imaging.

[43]  Léon Bottou,et al.  On-line learning and stochastic approximations , 1999 .

[44]  Mathews Jacob,et al.  MoDL: Model-Based Deep Learning Architecture for Inverse Problems , 2017, IEEE Transactions on Medical Imaging.

[45]  Reinhard Heckel,et al.  Deep Decoder: Concise Image Representations from Untrained Non-convolutional Networks , 2018, ICLR.

[46]  Elsa D. Angelini,et al.  An Unbiased Risk Estimator for Image Denoising in the Presence of Mixed Poisson–Gaussian Noise , 2014, IEEE Transactions on Image Processing.

[47]  Alessandro Foi,et al.  Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering , 2007, IEEE Transactions on Image Processing.

[48]  Jeffrey A. Fessler,et al.  Ieee Transactions on Image Processing: to Appear Hybrid Poisson/polynomial Objective Functions for Tomographic Image Reconstruction from Transmission Scans , 2022 .

[49]  Quanzheng Li,et al.  Iterative Low-Dose CT Reconstruction With Priors Trained by Artificial Neural Network , 2017, IEEE Transactions on Medical Imaging.

[50]  Se Young Chun,et al.  Training Deep Learning based Denoisers without Ground Truth Data , 2018, NeurIPS.

[51]  Hongming Shan,et al.  Dual Network Architecture for Few-view CT - Trained on ImageNet Data and Transferred for Medical Imaging , 2019, Developments in X-Ray Tomography XII.

[52]  Ciprian Catana,et al.  PET Image Reconstruction Using Deep Image Prior , 2019, IEEE Transactions on Medical Imaging.

[53]  Jeffrey A. Fessler,et al.  Fast X-Ray CT Image Reconstruction Using a Linearized Augmented Lagrangian Method With Ordered Subsets , 2014, IEEE Transactions on Medical Imaging.

[54]  Daniel Rueckert,et al.  A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction , 2017, IEEE Transactions on Medical Imaging.

[55]  Jong Chul Ye,et al.  A deep convolutional neural network using directional wavelets for low‐dose X‐ray CT reconstruction , 2016, Medical physics.

[56]  Jonathan T. Barron,et al.  Unprocessing Images for Learned Raw Denoising , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[57]  Y. Nesterov.  A method for solving the convex programming problem with convergence rate O(1/k^2) , 1983 .

[58]  Alessandro Foi,et al.  Optimal Inversion of the Anscombe Transformation in Low-Count Poisson Image Denoising , 2011, IEEE Transactions on Image Processing.

[59]  Timo Aila,et al.  Improved Self-Supervised Deep Image Denoising , 2019 .

[60]  Se Young Chun,et al.  Extending Stein's unbiased risk estimator to train deep denoisers with correlated pairs of noisy images , 2019, NeurIPS.