Advertisement

Inter-Scanner Harmonization of High Angular Resolution DW-MRI Using Null Space Deep Learning

  • Vishwesh NathEmail author
  • Prasanna Parvathaneni
  • Colin B. Hansen
  • Allison E. Hainline
  • Camilo Bermudez
  • Samuel Remedios
  • Justin A. Blaber
  • Kurt G. Schilling
  • Ilwoo Lyu
  • Vaibhav Janve
  • Yurui Gao
  • Iwona Stepniewska
  • Baxter P. Rogers
  • Allen T. Newton
  • L. Taylor Davis
  • Jeff Luci
  • Adam W. Anderson
  • Bennett A. Landman
Conference paper
Part of the Mathematics and Visualization book series (MATHVISUAL)

Abstract

Diffusion-weighted magnetic resonance imaging (DW-MRI) allows for non-invasive imaging of the local fiber architecture of the human brain at a millimetric scale. Multiple classical approaches have been proposed to detect both single (e.g., tensors) and multiple (e.g., constrained spherical deconvolution, CSD) fiber population orientations per voxel. However, existing techniques generally exhibit low reproducibility across MRI scanners. Herein, we propose a data-driven technique using a neural network design which exploits two categories of data. First, training data were acquired on three squirrel monkey brains using ex-vivo DW-MRI and histology of the brain. Second, repeated scans of human subjects were acquired on two different scanners to augment the learning of the network proposed. To use these data, we propose a new network architecture, the null space deep network (NSDN), to simultaneously learn on traditional observed/truth pairs (e.g., MRI-histology voxels) along with repeated observations without a known truth (e.g., scan-rescan MRI). The NSDN was tested on twenty percent of the histology voxels that were kept completely blind to the network. NSDN significantly improved absolute performance relative to histology by 3.87% over CSD and 1.42% over a recently proposed deep neural network approach. More-over, it improved reproducibility on the paired data by 21.19% over CSD and 10.09% over a recently proposed deep approach. Finally, NSDN improved generalizability of the model to a third in-vivo human scanner (which was not used in training) by 16.08% over CSD and 10.41% over a recently proposed deep learning approach. This work suggests that data-driven approaches for local fiber reconstruction are more reproducible, informative, precise and offer a novel, practical method for determining these models.

Keywords

Diffusion HARDI DW-MRI Null space CSD Harmonization Inter-scanner Deep learning Histology 

Notes

Acknowledgements

This work was supported by R01EB017230 (Landman), RO1NS058639 (Anderson), S10RR17799 (Gore) and T32EB001628. This work was conducted in part using the resources of the Advanced Computing Center for Research and Education at Vanderbilt University, Nashville, TN. This project was supported in part by the National Center for Research Resources, Grant UL1 RR024975-01, and is now at the National Center for Advancing Translational Sciences, Grant 2 UL1 TR000445-06. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. This work has been supported by Nvidia with supplement of hardware resources (GPUs). Glyph visualizations were supported using  [27].

References

  1. 1.
    Le Bihan, D.: Looking into the functional architecture of the brain with diffusion MRI. Nature Rev. Neurosci. 4(6), 469 (2003)CrossRefGoogle Scholar
  2. 2.
    Basser, P.J., Mattiello, J., LeBihan, D.: MR diffusion tensor spectroscopy and imaging. Biophys. J. 66(1), 259–267 (1994)CrossRefGoogle Scholar
  3. 3.
    Tournier, J., Calamante, F., Connelly, A.: How many diffusion gradient directions are required for HARDI. In: Proceedings of the International Society Magnetic Resonance in Medicine (2009)Google Scholar
  4. 4.
    Tournier, J.-D., Calamante, F., Connelly, A.: Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution. Neuroimage 35(4), 1459–1472 (2007)CrossRefGoogle Scholar
  5. 5.
    Tuch, D.S.: Qball imaging. Magn. Reson. Med. 52(6), 1358–1372 (2004)CrossRefGoogle Scholar
  6. 6.
    Anderson, A.W.: Measurement of fiber orientation distributions using high angular resolution diffusion imaging. Magn. Reson. Med. 54(5), 1194–1206 (2005)CrossRefGoogle Scholar
  7. 7.
    Jansons, K.M., Alexander, D.C.: Persistent angular structure: new insights from diffusion magnetic resonance imaging data. Inverse Probl. 19(5), 1031 (2003)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Gorczewski, K., Mang, S., Klose, U.: Reproducibility and consistency of evaluation techniques for HARDI data. Magn. Reson. Material. Phys. Biol. Med. 22(1), 63 (2009)CrossRefGoogle Scholar
  9. 9.
    Nath, V., et al.: Comparison of multi-fiber reproducibility of PAS-MRI and Q-ball with empirical multiple b-value HARDI. In: Medical Imaging 2017: Image Processing (2017). (International Society for Optics and Photonics)Google Scholar
  10. 10.
    Helmer, K., et al.: Multi-site study of diffusion metric variability: effects of site, vendor, field strength, and echo time on regions-of-interest and histogram-bin analyses. In: Medical Imaging 2016: Biomedical Applications in Molecular, Structural, and Functional Imaging. International Society for Optics and Photonics (2016)Google Scholar
  11. 11.
    Huo, J., et al.: Between-scanner and between-visit variation in normal white matter apparent diffusion coefficient values in the setting of a multi-center clinical trial. Clin. Neuroradiol. 26(4), 423–430 (2016)CrossRefGoogle Scholar
  12. 12.
    Schilling, K., et al.: Comparison of 3D orientation distribution functions measured with confocal microscopy and diffusion MRI. Neuroimage 129, 185–197 (2016)CrossRefGoogle Scholar
  13. 13.
    Mirzaalian, H., et al.: Harmonizing diffusion MRI data across multiple sites and scanners. In: International Conference on Medical Image Computing and Computer Assisted Intervention. Springer (2015)Google Scholar
  14. 14.
    Mirzaalian, H., et al.: Inter-site and inter-scanner diffusion MRI data harmonization. NeuroImage 135, 311–323 (2016)CrossRefGoogle Scholar
  15. 15.
    Fortin, J.-P., et al.: Harmonization of multi-site diffusion tensor imaging data. Neuroimage 161, 149–170 (2017)CrossRefGoogle Scholar
  16. 16.
    Nath, Vishwesh, et al.: Deep learning captures more accurate diffusion fiber orientations distributions than constrained spherical deconvolution. In: ISMRM 2018, Paris, FranceGoogle Scholar
  17. 17.
    Stolp, H., et al.: Voxel-wise comparisons of cellular microstructure and diffusion-MRI in mouse hippocampus using 3D bridging of optically-clear histology with neuroimaging data (3D-BOND). Scientific reports 8(1), 4011 (2018)CrossRefGoogle Scholar
  18. 18.
    Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. In: Advances in neural information processing systems (2012)Google Scholar
  19. 19.
    Andersson, J.L., Skare, S., Ashburner, J.: How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. Neuroimage 20(2), 870–888 (2003)CrossRefGoogle Scholar
  20. 20.
    Andersson, J.L., Sotiropoulos, S.N.: An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage 125, 1063–1078 (2016)CrossRefGoogle Scholar
  21. 21.
    Smith, S.M.: Fast robust automated brain extraction. Hum. Brain Mapp. 17(3), 143–155 (2002)CrossRefGoogle Scholar
  22. 22.
    Nath, Vishwesh, et al.,: SHARD: spherical harmonic-based robust outlier detection for HARDI methods. In: Medical Imaging 2018: Image Processing. Vol. 10574. International Society for Optics and Photonics (2018)Google Scholar
  23. 23.
    Zhang, Y., Brady, M., Smith, S.: Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans. Med. Imaging 20(1), 45–57 (2001)CrossRefGoogle Scholar
  24. 24.
    Li, S., et al.: A discriminative null space based deep learning approach for person re-identification. In: 2016 4th International Conference on Cloud Computing and Intelligence Systems (CCIS). IEEE (2016)Google Scholar
  25. 25.
    Descoteaux, M., et al.: Apparent diffusion coefficients from high angular resolution diffusion imaging: estimation and applications. Magn. Reson. Med. 56(2), 395–410 (2006)CrossRefGoogle Scholar
  26. 26.
    Hinton, G., Srivastava, N., Swersky, K.: Neural networks for machine learning-lecture 6a-overview of mini-batch gradient descent. Coursera Lecture slides (2012)Google Scholar
  27. 27.
    Blaber, J., Schilling, K., Landman, B.: (2018, March 6). justinblaberdwmri visualizer: First release of dwmri\(\_\)visualizer (Version v1.0.0). Zenodo. https://doi.org/10.5281/zenodo.1191107

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Vishwesh Nath
    • 1
    Email author
  • Prasanna Parvathaneni
    • 1
  • Colin B. Hansen
    • 1
  • Allison E. Hainline
    • 3
  • Camilo Bermudez
    • 2
  • Samuel Remedios
    • 4
  • Justin A. Blaber
    • 1
  • Kurt G. Schilling
    • 2
  • Ilwoo Lyu
    • 1
  • Vaibhav Janve
    • 6
  • Yurui Gao
    • 6
  • Iwona Stepniewska
    • 5
  • Baxter P. Rogers
    • 6
  • Allen T. Newton
    • 6
  • L. Taylor Davis
    • 7
  • Jeff Luci
    • 8
  • Adam W. Anderson
    • 2
    • 6
  • Bennett A. Landman
    • 1
    • 2
    • 6
  1. 1.EECSVanderbilt UniversityNashvilleUSA
  2. 2.BMEVanderbilt UniversityNashvilleUSA
  3. 3.Biostatistics, Vanderbilt UniversityNashvilleUSA
  4. 4.CS, Middle Tennessee State UniversityMurfressboroUSA
  5. 5.Psychology, Vanderbilt UniversityNashvilleUSA
  6. 6.VUIISVanderbilt UniversityNashvilleUSA
  7. 7.VUMCVanderbilt UniversityNashvilleUSA
  8. 8.BMEUniversity of Texas at AustinAustinUSA

Personalised recommendations