Abstract
Lung nodule classification based on low-dose computed tomography (LDCT) images has attracted major attention thanks to the reduced radiation dose and its potential for early diagnosis of lung cancer from LDCT-based lung cancer screening. However, LDCT images suffer from severe noise, largely influencing the performance of lung nodule classification. Current methods combining denoising and classification tasks typically require the corresponding normal-dose CT (NDCT) images as the supervision for the denoising task, which is impractical in the context of clinical diagnosis using LDCT. To jointly train these two tasks in a unified framework without the NDCT images, this paper introduces a novel self-supervised method, termed strided Noise2Neighbors or SN2N, for blind medical image denoising and lung nodule classification, where the supervision is generated from noisy input images. More specifically, the proposed SN2N can construct the supervision information from its neighbors for LDCT denoising, which does not need NDCT images anymore. The proposed SN2N method enables joint training of LDCT denoising and lung nodule classification tasks by using self-supervised loss for denoising and cross-entropy loss for classification. Extensively experimental results on the Mayo LDCT dataset demonstrate that our SN2N achieves competitive performance compared with the supervised learning methods that have paired NDCT images as supervision. Moreover, our results on the LIDC-IDRI dataset show that the joint training of LDCT denoising and lung nodule classification significantly improves the performance of LDCT-based lung nodule classification.
Similar content being viewed by others
Data Availability
The datasets of Mayo LDCT and LIDC-IDRI are publicly available at https://wiki.cancerimagingarchive.net/pages/viewpage.action?pageId=52758026 and https://wiki.cancerimagingarchive.net/display/Public/LIDC-IDRI, respectively.
Code Availability
Code is available at https://github.com/ymLeiFDU/Strided_Noise2Neighbors.
References
Aharon M, Elad M, Bruckstein A (2006) k-SVD: an algorithm for designing overcomplete dictionaries for sparse representation. IEEE Trans Signal Process 54(11):4311–4322. https://doi.org/10.1109/TSP.2006.881199
Armato SG III, McLennan G, Bidaut L et al (2011) The lung image database consortium (LIDC) and image database resource initiative (IDRI): a completed reference database of lung nodules on CT scans. Med Phys 38(2):915–931. https://doi.org/10.1118/1.3528204
Batson J, Royer L (2019): Noise2Self: Blind denoising by self-supervision. Proceedings of International Conference on Machine Learning (ICML) pp 524–533.
Beckham C, Pal C (2017) Unimodal probability distributions for deep ordinal classification. In: International Conference on Machine Learning (ICML), pp 411–419
Cao GT, Huang TT, Hou K, Cao WM, Liu P, Zhang JW (2018) 3D convolutional neural networks fusion model for lung nodule detection on clinical CT scans. In: IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp 973–978. IEEE. https://doi.org/10.1109/BIBM.2018.8621468
Chen H, Zhang Y, Kalra MK et al (2017) Low-dose CT with a residual encoder-decoder convolutional neural network. IEEE Trans Med Imaging 36(12):2524–2535. https://doi.org/10.1109/TMI.2017.2715284
Chen H, Zhang Y, Chen YJ et al (2018) LEARN: Learned experts’ assessment-based reconstruction network for sparse-data CT. IEEE Trans Med Imaging 37(6):1333–1347. https://doi.org/10.1109/TMI.2018.2805692
Feruglio PF, Vinegoni C, Gros J, Sbarbati A, Weissleder R (2010) Block matching 3D random noise filtering for absorption optical projection tomography. Phys Med Biol 55(18):5401. https://doi.org/10.1088/0031-9155/55/18/009
Hussein S, Gillies R, Cao K, Song Q, Bagci U (2017) TumorNet: Lung nodule characterization using multi-view convolutional neural network with Gaussian process. In: IEEE International Symposium on Biomedical Imaging (ISBI), pp 1007–1010. https://doi.org/10.1109/ISBI.2017.7950686
Janocha K, Czarnecki WM (2017) On loss functions for deep neural networks in classification. Schedae Informaticae 25:49–59. https://doi.org/10.4467/20838476SI.16.004.6185
Kingma DP, Ba J (2014) Adam: A method for stochastic optimization. In: International Conference on Learning Representations (ICLR)
Krull A, Buchholz TO, Jug F (2019) Noise2Void: learning denoising from single noisy images. In: Proceedings of the IEEE Conference Computer Vision Pattern Recognition (CVPR), pp 2129–2137. https://doi.org/10.1109/CVPR.2019.00223
Lehtinen J, Munkberg J, Hasselgren J et al (2018) Noise2Noise: Learning image restoration without clean data. Proceedings of International Conference on Machine Learning (ICML) pp 2965–2974
Lei YM, Zhu HP, Zhang JP, Shan HM (2020) Meta ordinal regression forest for learning with unsure lung nodules. In: 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp 442–445. https://doi.org/10.1109/BIBM49941.2020.9313554
Lei YM, Tian YK, Shan HM, Zhang JP, Wang G, Kalra MK et al (2020b) Shape and margin-aware lung nodule classification in low-dose CT images via soft activation mapping. Med Image Anal 60:101628. https://doi.org/10.1016/j.media.2019.101628
Lei YM, Shan HM, Zhang JP (2021) Meta ordinal weighting net for improving lung nodule classification. In: ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp 1210–1214. https://doi.org/10.1109/ICASSP39728.2021.9413622
Li ZB, Yu LF, Trzasko JD et al (2014) Adaptive nonlocal means filtering based on local noise level for CT denoising. Med Phys. https://doi.org/10.1118/1.4851635
Liu D, Wen BH, Liu XM, Wang ZY, Huang TS (2017) When image denoising meets high-level vision tasks: A deep learning approach. Proceedings of International Joint Conference Artificial Intelligence (IJCAI) pp 842–848. https://doi.org/10.24963/ijcai.2018/117
Liu XF, Zou Y, Song YH, Yang C, You J, Kumar BVKV (2018) Ordinal regression with neuron stick-breaking for medical diagnosis. In: Proceedings of the European Conference on Computer Vision (ECCV) Workshops, pp 335–344. https://doi.org/10.1007/978-3-030-11024-6_23
Liu D, Cheng BW, Wang ZY, Zhang HC, Huang TS (2019) Enhance visual recognition under adverse conditions via deep networks. IEEE Trans Image Process 28(9):4401–4412. https://doi.org/10.1109/TIP.2019.2908802
Liu D, Wen BH, Jiao JB, Liu XM, Wang ZY, Huang TS (2020) Connecting image denoising and high-level vision tasks via deep learning. IEEE Trans Image Process 29:3695–3706. https://doi.org/10.1109/TIP.2020.2964518
Nguyen A, Yosinski J, Clune J (2015) Deep neural networks are easily fooled: High confidence predictions for unrecognizable images. In: Proceedings of the IEEE Conference Computer Vision Pattern Recognition (CVPR), pp 427–436. https://doi.org/10.1109/CVPR.2015.7298640
Ronneberger O, Fischer P, Brox T (2015) U-Net: Convolutional networks for biomedical image segmentation. In: International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp 234–241. https://doi.org/10.1007/978-3-319-24574-4_28
Setio AAA, Ciompi F, Litjens G et al (2016) Pulmonary nodule detection in CT images: false positive reduction using multi-view convolutional networks. IEEE Trans Med Imaging 35(5):1160–1169. https://doi.org/10.1109/TMI.2016.2536809
Shan HM, Wang G, Kalra MK, de Souza R, Zhang JP (2017) Enhancing transferability of features from pretrained deep neural networks for lung nodule classification. In: Proceedings of the 2017 International Conference on Fully Three-Dimensional Image Reconstruction in Radiology and Nuclear Medicine (Fully3D), pp 65–68.
Shan HM, Zhang Y, Yang QS et al (2018) 3-D convolutional encoder-decoder network for low-dose CT via transfer learning from a 2-D trained network. IEEE Trans Med Imaging 37(6):1522–1534. https://doi.org/10.1109/TMI.2018.2832217
Shan HM, Jia X, Yan PK, Li YY, Paganetti H, Wang G (2020) Synergizing medical imaging and radiotherapy with deep learning. Mach Learn Sci Technol. https://doi.org/10.1088/2632-2153/ab869f
Shen W, Zhou M, Yang F, Yang CY, Tian J (2015) Multi-scale convolutional neural networks for lung nodule classification. In: International Conference on Information Processing in Medical Imaging (IPMI), pp. 588–599. https://doi.org/10.1007/978-3-319-19992-4_46
Shin HC, Roth HR, Gao MC et al (2016) Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE Trans Med Imaging 35(5):1285–1298. https://doi.org/10.1109/TMI.2016.2528162
Simonyan K, Zisserman A (2015) Very deep convolutional networks for large-scale image recognition. In: International Conference on Learning Representations (ICLR)
Szegedy C, Zaremba W, Sutskever L et al (2014) Intriguing properties of neural networks. In: International Conference Learning Representations (ICLR).
Wang ZY, Chang SY, Yang YZ, Liu D, Huang TS (2016) Studying very low resolution recognition using deep networks. In: Proceedings of the IEEE Conference on Computer Vision Pattern Recognition (CVPR), pp. 4792–4800. https://doi.org/10.1109/CVPR.2016.518
Woolson, RF (2007) Wilcoxon signed-rank test. Wiley encyclopedia of clinical trials. pp 1–3. https://doi.org/10.1201/9780429294402-204
Wu DF, Kim K, Fakhri GE, Li QZ (2017) A cascaded convolutional neural network for x-ray low-dose CT image denoising. arXiv:1705.04267
Wu JQ, Timofte R, Huang Z, Gool L (2017) On the relation between color image denoising and classification. arXiv:1704.01372
Wu, BT, Sun XW, Hu LJ, Wang YZ (2019) Learning with unsure data for medical image diagnosis. In: IEEE International Conference on Computer Vision. (ICCV), pp 10590–10599. https://doi.org/10.1109/ICCV.2019.01069
Xia WJ, Lu ZX, Huang YQ et al (2021a) CT reconstruction with PDF: parameter-dependent framework for data from multiple geometries and dose levels. IEEE Trans Med Imaging. https://doi.org/10.1109/TMI.2021.3085839
Xia WJ, Lu ZX, Huang YQ et al (2021b) MAGIC: Manifold and graph integrative convolutional network for low-dose CT reconstruction. IEEE Trans Med Imaging. https://doi.org/10.1109/TMI.2021.3088344
Zhang K, Zuo WM, Chen YJ, Meng DY, Zhang L (2017) Beyond a Gaussian denoiser: residual learning of deep CNN for image denoising. IEEE Trans Image Process 26(7):3142–3155. https://doi.org/10.1109/TIP.2017.2662206
Zhang JJ, Chao HQ, Xu XN, Niu C, Wang G, Yan PK (2021) Task-oriented low-dose CT image denoising. arXiv:2103.13557
Funding
This work was supported in part by National Natural Science Foundation of China (No. 62101136), Shanghai Municipal of Science and Technology Project (No. 20JC1419500), Shanghai Sailing Program (No. 21YF1402800), the Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX01), ZJLab, Shanghai Center for Brain Science and Brain-Inspired Technology, the National Key R&D Program of China (No. 2018YFB1305104), and the Natural Science Foundation of Shanghai (No. 21ZR1403600).
Author information
Authors and Affiliations
Contributions
(I) Conception and design: All authors; (II) Collection of data: YL, HS; (III) Analysis of results: All authors; (IV) Manuscript writing and approval: All authors.
Corresponding author
Ethics declarations
Conflicts of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval
Not applicable.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Rights and permissions
About this article
Cite this article
Lei, Y., Zhang, J. & Shan, H. Strided Self-Supervised Low-Dose CT Denoising for Lung Nodule Classification. Phenomics 1, 257–268 (2021). https://doi.org/10.1007/s43657-021-00025-y
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s43657-021-00025-y