DOI QR코드

DOI QR Code

Exploring Generalization Capacity of Artificial Neural Network for Myelin Water Imaging

  • Lee, Jieun (Department of Electrical and Computer Engineering, Seoul National University) ;
  • Choi, Joon Yul (Department of Electrical and Computer Engineering, Seoul National University) ;
  • Shin, Dongmyung (Department of Electrical and Computer Engineering, Seoul National University) ;
  • Kim, Eung Yeop (Department of Radiology, Gil Medical Center, Gachon University College of Medicine) ;
  • Oh, Se-Hong (Division of Biomedical Engineering, Hankuk University of Foreign Studies) ;
  • Lee, Jongho (Department of Electrical and Computer Engineering, Seoul National University)
  • Received : 2020.04.10
  • Accepted : 2020.06.12
  • Published : 2020.12.31

Abstract

Purpose: To understand the effects of datasets with various parameters on pre-trained network performance, the generalization capacity of the artificial neural network for myelin water imaging (ANN-MWI) is explored by testing datasets with various scan protocols (i.e., resolution and refocusing RF pulse shape) and types of disorders (i.e., neuromyelitis optica and edema). Materials and Methods: ANN-MWI was trained to generate a T2 distribution, from which the myelin water fraction value was measured. The training and test datasets were acquired from healthy controls and multiple sclerosis patients using a multi-echo gradient and spin-echo sequence with the same scan protocols. To test the generalization capacity of ANN-MWI, datasets with different settings were utilized. The datasets were acquired or generated with different resolutions, refocusing pulse shape, and types of disorders. For all datasets, the evaluation was performed in a white matter mask by calculating the normalized root-mean-squared error (NRMSE) between the results from the conventional method and ANN-MWI. Additionally, for the patient datasets, the NRMSE was calculated in each lesion mask. Results: The results of ANN-MWI showed high reliability in generating myelin water fraction maps from the datasets with different resolutions. However, the increased errors were reported for the datasets with different refocusing pulse shapes and disorder types. Specifically, the region of lesions in edema patients reported high NRMSEs. These increased errors indicate the dependency of ANN-MWI on refocusing pulse flip angles and T2 characteristics. Conclusion: This study proposes information about the generalization accuracy of a trained network when applying deep learning to processing myelin water imaging.

Keywords

References

  1. Yoon J, Gong E, Chatnuntawech I, et al. Quantitative susceptibility mapping using deep neural network: QSMnet. Neuroimage 2018;179:199-206 https://doi.org/10.1016/j.neuroimage.2018.06.030
  2. Cohen O, Zhu B, Rosen MS. MR fingerprinting Deep RecOnstruction NEtwork (DRONE). Magn Reson Med 2018;80:885-894 https://doi.org/10.1002/mrm.27198
  3. Kwon K, Kim D, Park H. A parallel MR imaging method using multilayer perceptron. Med Phys 2017;44:6209-6224 https://doi.org/10.1002/mp.12600
  4. Lee D, Lee J, Ko J, Yoon J, Ryu K, Nam Y. Deep learning in MR image processing. Investig Magn Reson Imaging 2019;23:81-99 https://doi.org/10.13104/imri.2019.23.2.81
  5. Ben-David S, Blitzer J, Crammer K, Pereira F. Analysis of representations for domain adaptation. Advances in Neural Information Processing Systems (NIPS) 2006:137-144
  6. Martensson G, Ferreira D, Granberg T, et al. The reliability of a deep learning model in clinical out-of-distribution MRI data: a multicohort study. Med Image Anal 2020;66:101714
  7. Jung W, Yoon J, Ji S, et al. Exploring linearity of deep neural network trained QSM: QSMnet. Neuroimage 2020;211:116619 https://doi.org/10.1016/j.neuroimage.2020.116619
  8. Knoll F, Hammernik K, Kobler E, Pock T, Recht MP, Sodickson DK. Assessment of the generalization of learned image reconstruction and the potential for transfer learning. Magn Reson Med 2019;81:116-128 https://doi.org/10.1002/mrm.27355
  9. Zhang C, Bengio S, Hardt M, Recht B, Vinyals O. Understanding deep learning requires rethinking generalization. arXiv preprint arXiv:1611.03530, 2016
  10. Youn SW, Kwon OD, Hwang MJ. Multi-parametric quantitative MRI for measuring myelin loss in hyperglycemia-induced hemichorea. Investig Magn Reson Imaging 2019;23:148-156 https://doi.org/10.13104/imri.2019.23.2.148
  11. Seo JP, Kwon YH, Jang SH. Mini-review of studies reporting the repeatability and reproducibility of diffusion tensor imaging. Investig Magn Reson Imaging 2019;23:26-33 https://doi.org/10.13104/imri.2019.23.1.26
  12. Lee J, Lee D, Choi JY, Shin D, Shin HG, Lee J. Artificial neural network for myelin water imaging. Magn Reson Med 2020;83:1875-1883 https://doi.org/10.1002/mrm.28038
  13. MacKay A, Whittall K, Adler J, Li D, Paty D, Graeb D. In vivo visualization of myelin water in brain by magnetic resonance. Magn Reson Med 1994;31:673-677 https://doi.org/10.1002/mrm.1910310614
  14. Prasloski T, Madler B, Xiang QS, MacKay A, Jones C. Applications of stimulated echo correction to multicomponent T2 analysis. Magn Reson Med 2012;67:1803-1814 https://doi.org/10.1002/mrm.23157
  15. Prasloski T, Rauscher A, MacKay AL, et al. Rapid whole cerebrum myelin water imaging using a 3D GRASE sequence. Neuroimage 2012;63:533-539 https://doi.org/10.1016/j.neuroimage.2012.06.064
  16. Choi JY, Jeong IH, Oh SH, et al. Evaluation of normalappearing white matter in multiple sclerosis using direct visualization of short transverse relaxation time component (ViSTa) myelin water imaging and gradient echo and spin echo (GRASE) myelin water imaging. J Magn Reson Imaging 2019;49:1091-1098 https://doi.org/10.1002/jmri.26278
  17. Jeong IH, Choi JY, Kim SH, et al. Comparison of myelin water fraction values in periventricular white matter lesions between multiple sclerosis and neuromyelitis optica spectrum disorder. Mult Scler 2016;22:1616-1620 https://doi.org/10.1177/1352458516636247
  18. Pauly J, Le Roux P, Nishimura D, Macovski A. Parameter relations for the Shinnar-Le Roux selective excitation pulse design algorithm [NMR imaging]. IEEE Trans Med Imaging 1991;10:53-65 https://doi.org/10.1109/42.75611
  19. Jeong IH, Choi JY, Kim SH, et al. Normal-appearing white matter demyelination in neuromyelitis optica spectrum disorder. Eur J Neurol 2017;24:652-658 https://doi.org/10.1111/ene.13266
  20. Borich MR, Mackay AL, Vavasour IM, Rauscher A, Boyd LA. Evaluation of white matter myelin water fraction in chronic stroke. Neuroimage Clin 2013;2:569-580 https://doi.org/10.1016/j.nicl.2013.04.006
  21. Dvorak AV, Ljungberg E, Vavasour IM, et al. Rapid myelin water imaging for the assessment of cervical spinal cord myelin damage. Neuroimage Clin 2019;23:101896 https://doi.org/10.1016/j.nicl.2019.101896
  22. Baumeister TR, Kim JL, Zhu M, McKeown MJ. White matter myelin profiles linked to clinical subtypes of Parkinson's disease. J Magn Reson Imaging 2019;50:164-174 https://doi.org/10.1002/jmri.26543
  23. Lakhani B, Hayward KS, Boyd LA. Hemispheric asymmetry in myelin after stroke is related to motor impairment and function. Neuroimage Clin 2017;14:344-353 https://doi.org/10.1016/j.nicl.2017.01.009
  24. Maas AL, Hannun AY, Ng AY. Rectifier nonlinearities improve neural network acoustic models. In Proc of the 30th Int Conf Mach Learn (ICML), Atlanta, GA, USA, 2013:3
  25. Kingma DP, Ba J. Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014
  26. Wilson DR, Martinez TR. The need for small learning rates on large problems. In Proc 2001 Int Joint Conf Neural Netw (IJCNN), Washington DC, USA, 2001:115-119
  27. Smith SL, Kindermans P-J, Ying C, Le QV. Don't decay the learning rate, increase the batch size. arXiv preprint arXiv:1711.00489, 2017
  28. Du YP, Chu R, Hwang D, et al. Fast multislice mapping of the myelin water fraction using multicompartment analysis of T2* decay at 3T: a preliminary postmortem study. Magn Reson Med 2007;58:865-870 https://doi.org/10.1002/mrm.21409
  29. Meyers SM, Laule C, Vavasour IM, et al. Reproducibility of myelin water fraction analysis: a comparison of region of interest and voxel-based analysis methods. Magn Reson Imaging 2009;27:1096-1103 https://doi.org/10.1016/j.mri.2009.02.001
  30. Smith SM. Fast robust automated brain extraction. Hum Brain Mapp 2002;17:143-155 https://doi.org/10.1002/hbm.10062
  31. Hennig J. Multiecho imaging sequences with low refocusing flip angles. J Magn Reson 1988;78:397-407 https://doi.org/10.1016/0022-2364(88)90128-X
  32. Wang J, Mao W, Qiu M, Smith MB, Constable RT. Factors influencing flip angle mapping in MRI: RF pulse shape, slice-select gradients, off-resonance excitation, and B0 inhomogeneities. Magn Reson Med 2006;56:463-468 https://doi.org/10.1002/mrm.20947
  33. Pan SJ, Yang Q. A survey on transfer learning. IEEE Trans Knowl Data Eng 2009;22:1345-1359 https://doi.org/10.1109/TKDE.2009.191
  34. He H, Garcia EA. Learning from imbalanced data. IEEE Trans Knowl Data Eng 2009;21:1263-1284 https://doi.org/10.1109/TKDE.2008.239
  35. Krawczyk B. Learning from imbalanced data: open challenges and future directions. Artificial Intelligence 2016;5:221-232
  36. Lee LE, Ljungberg E, Shin D, et al. Inter-vendor reproducibility of myelin water imaging using a 3D gradient and spin echo sequence. Front Neurosci 2018;12:854 https://doi.org/10.3389/fnins.2018.00854