Supervised denoising of diffusion-weighted magnetic resonance images using a convolutional neural network and transfer learning

• Autorzy: Jurek Jakub , Materka Andrzej , Ludwisiak Kamil , Majos Agata , Gorczewski Kamil , Cepuch Kamil , Zawadzka Agata

Identyfikatory

DOI

10.1016/j.bbe.2022.12.006

• Języki publikacji: EN

Abstrakty

EN

In this paper, we propose a method for reducing thermal noise in diffusion-weighted magnetic resonance images (DWI MRI) of the brain using a convolutional neural network (CNN) trained on realistic, synthetic MR data. Two reference methods are considered: a) averaging of repeated scans, a widespread method used in clinics to improve signal-to-noise ratio of MR images and b) the blockwise Non-Local Means (NLM) filter, one of the post-processing methods frequently used in DWI denoising. To obtain training data for transfer learning, the effects of echo-planar imaging (EPI) – Nyquist ghosting and ramp sampling – are modelled in a data-driven fashion. These effects are introduced to the digital phantom of brain anatomy (BrainWeb). Real noise maps are obtained from the MRI scanner with a brainDWI-designed protocol and later combined with simulated, noise-free EPI images. The Point Spread Function is measured in a DW image of an AJR-approved geometrical phantom. Inter-scan patient movement is captured from a brain scan of a healthy volunteer using image registration. The denoising methods are applied to the simulated EPI brain images and in real EPI DWI of the brain. The quality of denoised images is evaluated at several signal-to-noise ratios. The characteristics of noise residuals are studied thoroughly. A diffusion phantom is used to investigate the influence of denoising on ADC measurements. The method is also evaluated on a GRAPPA dataset. We show that our method outperforms NLM and image averaging and allows for a significant reduction in scan time by lowering the number of repeated scans. We also analyse the trained CNN denoisers and point out the challenges accompanying this denoising method.

Słowa kluczowe

EN

diffusion weighted imaging; DWI denoising; convolutional neural network; transfer learning; diffusion phantom; modelling of echo planar images

PL

obrazowanie dyfuzyjne; sieć neuronowa konwolucyjna; transfer wiedzy; fantom dyfuzyjny

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2023
• Tom: Vol. 43, no. 1
• Strony: 206--232
• Opis fizyczny: Bibliogr. 33 poz., rys., tab., wykr.

Twórcy

autor

Jurek Jakub

• Institute of Electronics, Lodz University of Technology, Aleja Politechniki 10, PL-93590 Lodz, Poland

autor

Materka Andrzej

• Institute of Electronics, Lodz University of Technology, Lodz, Poland

autor

Ludwisiak Kamil

• Department of Diagnostic Imaging, Independent Public Health Care, Central Clinical Hospital, Medical University of Lodz, Lodz, Poland

autor

Majos Agata

• Department of Radiological and Isotopic Diagnosis and Therapy, Medical University of Lodz, Lodz, Poland

autor

Gorczewski Kamil

• Siemens Healthcare GmbH, Erlangen, Germany

autor

Cepuch Kamil

• Siemens Healthcare GmbH, Erlangen, Germany

autor

Zawadzka Agata

• Siemens Healthcare GmbH, Erlangen, Germany

Bibliografia

• [1] Lutsep HL, Albers GW, Decrespigny A, Kamat GN, Marks MP, Moseley ME. Clinical utility of diffusion-weighted magnetic resonance imaging in the assessment of ischemic stroke. Ann Neurol 1997;41(5):574-80. https://doi.org/10.1002/ ana.410410505.
• [2] Messina C, Bignone R, Bruno A, Bruno A, Bruno F, Calandri M, et al. D.A. and Diffusion-weighted imaging in oncology: An update. Cancers 2020;12(6):1493. https://doi.org/ 10.3390/cancers12061493.
• [3] Mansfield P. Multi-planar image formation using NMR spin echoes. J Phys C: Solid State Phys 1977;10(3):L55-8. https://doi. Org/10.1088/0022-3719/10/3/004.
• [4] Johansen-Berg H, Behrens TEJ, editors. Diffusion MRI: From Quantitative Measurement to In vivo Neuroanatomy. Elsevier; 2014. https://doi.org/10.1016/c2011-0-07047-3.
• [5] Brown RW, Cheng Y-CN, Haacke EM, Thompson MR, Venkatesan R. Magnetic resonance imaging: physical principles and sequence design. 2nd ed. John Wiley and Sons; 2014.
• [6] Himmelberg MM, Kurzawski JW, Benson NC, Pelli DG, Carrasco M, Winawer J. Cross-dataset reproducibility of human retinotopic maps. NeuroImage 2021;244:118609. https://doi.org/10.1016/j.neuroimage.2021.118609.
• [7] Tax CM, Bastiani M, Veraart J, Garyfallidis E, Irfanoglu MO. What’s new and what’s next in diffusion MRI preprocessing. NeuroImage 2022;249:118830. https://doi.org/10.1016/j. Neuroimage.2021.118830.
• [8] Derek P, Jones K, editors. Diffusion MRI. Oxford University Press; 2012. https://doi.org/10.1093/med/ 9780195369779.001.0001.
• [9] Gudbjartsson H, Patz S. The rician distribution of noisy MRI data. Magn Reson Med 1995;34:910-4. https://doi.org/10.1002/ mrm.1910340618.
• [10] Buades A, Coll B, Morel J. A non-local algorithm for image denoising. In: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2005), 20-26 June 2005. EEE Computer Society: San Diego, CA, USA; 2005. p. 60-5. https://doi.org/10.1109/CVPR.2005.38.
• [11] Coupe P, Yger P, Prima S, Hellier P, Kervrann C, Barillot C. An optimized blockwise nonlocal means denoising filter for 3-d magnetic resonance images. IEEE Trans Med Imaging 2008;27 (4):425-41. https://doi.org/10.1109/TMI.2007.906087.
• [12] Veraart J, Novikov DS, Christiaens D, Ades-aron B, Sijbers J, Fieremans E. Denoising of diffusion MRI using random matrix theory. NeuroImage 2016;142:394-406. https://doi.org/ 10.1016/j.neuroimage.2016.08.016.
• [13] Moeller S, Pisharady PK, Ramanna S, Lenglet C, Wu X, Dowdle L, Yacoub E, et al. Noise reduction with Distribution corrected (NORDIC) PCA in dMRI with complex-valued parameter-free locally low-rank processing. NeuroImage 2021;226:117539. https://doi.org/10.1016/j. Neuroimage.2020.117539.
• [14] Fadnavis S, Batson J, Garyfallidis E. Patch2self: Denoising diffusion MRI with self-supervised learning. In: Thirty-fourth Conference on Neural Information Processing Systems; 2020. arXiv:2011.01355.
• [15] Tian Q, Li Z, Fan Q, Polimeni JR, Bilgic B, Salat DH, et al. SDnDTI: Self-supervised deep learning-based denoising for diffusion tensor MRI. NeuroImage 2022;253:119033. https:// doi.org/10.1016/j.neuroimage.2022.119033.
• [16] Batson J, Royer L. Noise2self: Blind denoising by selfsupervision. In: Chaudhuri K, Salakhutdinov R. editors. Proceedings of the 36th International Conference on Machine Learning, ICML 2019, 9-15 June 2019, Long Beach, California, USA, Vol. 97 of Proceedings of Machine Learning Research, PMLR; 2019. p. 524-33. http://proceedings.mlr.press/ v97/batson19a.html.
• [17] Lehtinen J, Munkberg J, Hasselgren J, Laine S, Karras T, Aittala M, Aila T. Noise2noise: Learning image restoration without clean data. In: Dy JG, Krause A. editors. Proceedings of the 35th International Conference on Machine Learning, ICML 2018, Stockholmsmässan, Stockholm, Sweden, July 10-15, 2018, Vol. 80 of Proceedings of Machine Learning Research, PMLR; 2018. p. 2971-80. http://proceedings.mlr.press/v80/ lehtinen18a.html.
• [18] Muckley MJ, Ades-Aron B, Papaioannou A, Lemberskiy G, Solomon E, Lui YW, et al. Training a neural network for gibbs and noise removal in diffusion MRI. Magn Reson Med 2020;85 (1):413-28. https://doi.org/10.1002/mrm.28395.
• [19] [link]. http://www.bic.mni.mcgill.ca/brainweb/.
• [20] Cocosco CA, Kollokian V, Kwan RK-S, Pike GB, Evans AC. Brainweb: Online interface to a 3D MRI simulated brain database. NeuroImage 1997;5:425.
• [21] Collins D, Zijdenbos A, Kollokian V, Sled J, Kabani N, Holmes C, et al. Design and construction of a realistic digital brain phantom. IEEE Trans Med Imaging 1998;17(3):463-8. https:// doi.org/10.1109/42.712135.
• [22] Kwan RKS, Evans AC, Pike GB. An extensible MRI simulator for post-processing evaluation. In: Lecture Notes in Computer Science. Berlin Heidelberg: Springer; 1996. p. 135-40. https://doi.org/10.1007/bfb0046947.
• [23] Kwan R-S, Evans A, Pike G. MRI simulation-based evaluation of image-processing and classification methods. IEEE Trans Med Imaging 1999;18:1085-97. https://doi.org/10.1109/ 42.816072.
• [24] Pizzolato M, Gilbert G, Thiran J-P, Descoteaux M, Deriche R. Adaptive phase correction of diffusion-weighted images. NeuroImage 2020;206:116274. https://doi.org/10.1016/j. Neuroimage.2019.116274.
• [25] D.C. Hansen. [link]. https://github.com/ismrmrd/ismrmrdpython-tools.
• [26] Schmitt F, Stehling MK, Turner R. Echo-Planar Imaging. Springer Berlin Heidelberg; 1998. https://doi.org/10.1007/978-3-642-80443-4.
• [27] Materka A, Górnisiewicz W, Kukula M. Elementy cyfrowego przetwarzania i analizy obrazów. Polskie Wydawnictwo Naukowe 1991.
• [28] Kellman P, McVeigh ER. Image reconstruction in snr units: A general method for snr measurement. Magn Reson Med 2005;54:1439-47. https://doi.org/10.1002/mrm.20713.
• [29] Dong C, Loy CC, He K, Tang X. Image super-resolution using deep convolutional networks. IEEE Trans Pattern Anal Mach Intell 2016;38:295-307. https://doi.org/10.1109/ TPAMI.2015.2439281.
• [30] Jurek J, Kociński M, Materka A, Elgalal M, Majos A. CNN-based superresolution reconstruction of 3d MR images using thickslice scans. Biocybernet Biomed Eng 2020;40(1):111-25. https://doi.org/10.1016/j.bbe.2019.10.003.
• [31] Zhang K, Zuo W, Chen Y, Meng D, Zhang L. Beyond a gaussian denoiser: Residual learning of deep CNN for image denoising. IEEE Trans Image Process 2017;26(7):3142-55. https://doi.org/10.1109/TIP.2017.2662206.
• [32] Wang Z, Simoncelli EP, Bovik AC. Multiscale structural similarity for image quality assessment. In: The Thirty-Seventh Asilomar Conference on Signals, Systems & Computers, vol. 2, IEEE; 2003. p. 1398-402. https://doi.org/10. 1109/acssc.2003.1292216.
• [33] Garyfallidis E, Brett M, Amirbekian B, Rokem A, Van Der Walt S, Descoteaux M, Nimmo-Smith I. Dipy, a library for the analysis of diffusion MRI data, Front Neuroinformat 2014;8. https://doi.org/10.3389/fninf.2014.00008. https://www. frontiersin.org/articles/10.3389/fninf.2014.00008.

Uwagi

PL

Opracowanie rekordu ze środków MEiN, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2022-2023)