Circular Non-uniform Sampling Patch Inputs for CNN Applied to Multiple Sclerosis Lesion Segmentation

  • Gustavo UlloaEmail author
  • Rodrigo Naranjo
  • Héctor Allende-Cid
  • Steren Chabert
  • Héctor Allende
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11401)


Convolutional Neural Networks (CNN) have been obtaining successful results in the task of image segmentation in recent years. These methods use as input the sampling obtained using square uniform patches centered on each voxel of the image, which could not be the optimal approach since there is a very limited use of global context. In this work we present a new construction method for the patches by means of a circular non-uniform sampling of the neighborhood of the voxels. This allows a greater global context with a radial extension with respect to the central voxel. This approach was applied on the 2015 Longitudinal MS Lesion Segmentation Challenge dataset, obtaining better results than approaches using square uniform and non-uniform patches with the same computational cost of the CNN models.


Convolutional Neural Networks Image segmentation Multiple sclerosis lesions Magnetic resonance imaging Non-uniform patch 



This work was supported by the Fondecyt Grant 1170123 and in part by Fondecyt Grant FB0821. Héctor Allende-Cid is supported by project Fondecyt Initiation into Research 11150248. Steren Chabert is supported by Centro de Investigación y Desarrollo en Ingeniería en Salud (CINGS).

The research of Gustavo Ulloa is also supported by the Incentive program for scientific initiation (PIIC) of the Universidad Técnica Federico Santa María DGIIP.


  1. 1.
    Birenbaum, A., Greenspan, H.: Multi-view longitudinal CNN for multiple sclerosis lesion segmentation. Eng. Appl. Artif. Intell. 65, 111–118 (2017)CrossRefGoogle Scholar
  2. 2.
    Carass, A., et al.: Longitudinal multiple sclerosis lesion segmentation: resource and challenge. NeuroImage 148, 77–102 (2017)CrossRefGoogle Scholar
  3. 3.
    Chollet, F., et al.: Keras (2015).
  4. 4.
    García-Lorenzo, D., Francis, S., Narayanan, S., Arnold, D.L., Collins, D.L.: Review of automatic segmentation methods of multiple sclerosis white matter lesions on conventional magnetic resonance imaging. Med. Image Anal. 17(1), 1–18 (2013)CrossRefGoogle Scholar
  5. 5.
    Ghafoorian, M., et al.: Non-uniform patch sampling with deep convolutional neural networks for white matter hyperintensity segmentation. In: 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI), pp. 1414–1417, April 2016Google Scholar
  6. 6.
    González-Villà, S., Oliver, A., Valverde, S., Wang, L., Zwiggelaar, R., Lladó, X.: A review on brain structures segmentation in magnetic resonance imaging. Artif. Intell. Med. 73, 45–69 (2016)CrossRefGoogle Scholar
  7. 7.
    Inglese, M., Oesingmann, N., Casaccia, P., Fleysher, L.: Progressive multiple sclerosis and gray matter pathology: an MRI perspective. Mt. Sinai J. Med.: J. Transl. Personalized Med. 78(2), 258–267 (2011)CrossRefGoogle Scholar
  8. 8.
    Mazziotta, J., et al.: A probabilistic atlas and reference system for the human brain: International consortium for brain mapping (ICBM). Philos. Trans. Royal Soc. London B: Biol. Sci. 356(1412), 1293–1322 (2001)CrossRefGoogle Scholar
  9. 9.
    Roy, S., Butman, J.A., Reich, D.S., Calabresi, P.A., Pham, D.L.: Multiple sclerosis lesion segmentation from brain MRI via fully convolutional neural networks. CoRR abs/1803.09172 (2018)Google Scholar
  10. 10.
    Valverde, S., et al.: Improving automated multiple sclerosis lesion segmentation with a cascaded 3D convolutional neural network approach. NeuroImage 155, 159–168 (2017)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Gustavo Ulloa
    • 1
    Email author
  • Rodrigo Naranjo
    • 1
  • Héctor Allende-Cid
    • 2
  • Steren Chabert
    • 3
  • Héctor Allende
    • 1
  1. 1.Universidad Técnica Federico Santa MaríaValparaísoChile
  2. 2.Pontificia Universidad Católica de ValparaísoValparaísoChile
  3. 3.CINGS, Universidad de ValparaísoValparaísoChile

Personalised recommendations