Longitudinal Multiple Sclerosis Lesion Segmentation Using 3d Convolutional Neural Networks

Abstract

We present our entry for the Longitudinal Multiple Sclerosis Challenge 2015 using 3D convolutional neural networks (CNN). We model a voxel-wise classifier using multi-channel 3D patches of MRI volumes as input. For each ground truth, a CNN is trained and the final segmentation is obtained by combining the probability outputs of these CNNs. Efficient training is achieved by using sub-sampling methods and sparse convolutions. We obtain accurate results with dice scores comparable to the inter-rater variability.

Extracted Key Phrases

3 Figures and Tables

Showing 1-7 of 7 references

Multi-modal brain tumor segmentation using deep convolutional neural networks

  • G Urban, M Bendszus, F A Hamprecht, J Kleesiek
  • 2014
1 Excerpt

Advanced normalization tools (ants)

  • B Brian, Nick Avants, Gang Tustison, Song
  • 2009
1 Excerpt

ANTs/ANTsR Brain Templates

  • Brian Avants, Nick Tustison
1 Excerpt