Left Atrial Segmentation In a Few Seconds Using Fully Convolutional Network and Transfer Learning

From LRDE

Abstract

In this paper, we propose a fast automatic method that segments left atrial cavity from 3D GE-MRIs without any manual assistance, using a fully convolutional network (FCN) and transfer learning. This FCN is the base network of VGG-16, pre-trained on ImageNet for natural image classification, and fine tuned with the training dataset of the MICCAI 2018 Atrial Segmentation Challenge. It relies on the "pseudo-3D" method published at ICIP 2017, which allows for segmenting objects from 2D color images which contain 3D information of MRI volumes. For each slice of the volume to segment, we consider three imagescorresponding to the , and slices of the original volume. These three gray-level 2D images are assembled to form a 2D RGB color image (one image per channel). This image is the input of the FCN to obtain a 2D segmentation of the slice. We process all slices, then stack the results to form the 3D output segmentation. With such a technique, the segmentation of the left atrial cavity on a 3D volume takes only a few seconds. We obtain a Dice score of 0.92 both on the training set in our experiments before the challenge, and on the test set of the challenge.

Documents

Bibtex (lrde.bib)

@InProceedings{	  puybareau.18.stacom,
  author	= {\'Elodie Puybareau and Zhou Zhao and Younes Khoudli and
		  Edwin Carlinet and Yongchao Xu and J\'er\^ome Lacotte and
		  Thierry G\'eraud},
  title		= {Left Atrial Segmentation In a Few Seconds Using Fully
		  Convolutional Network and Transfer Learning},
  booktitle	= {Proceedings of the Workshop on Statistical Atlases and
		  Computational Modelling of the Heart (STACOM 2018), in
		  conjunction with MICCAI},
  year		= 2019,
  series	= {Lecture Notes in Computer Science},
  publisher	= {Springer},
  volume	= {11395},
  pages		= {339--347},
  doi		= {10.1007/978-3-030-12029-0_37},
  abstract	= {In this paper, we propose a fast automatic method that
		  segments left atrial cavity from 3D GE-MRIs without any
		  manual assistance, using a fully convolutional network
		  (FCN) and transfer learning. This FCN is the base network
		  of VGG-16, pre-trained on ImageNet for natural image
		  classification, and fine tuned with the training dataset of
		  the MICCAI 2018 Atrial Segmentation Challenge. It relies on
		  the "pseudo-3D" method published at ICIP 2017, which allows
		  for segmenting objects from 2D color images which contain
		  3D information of MRI volumes. For each $n^{\text{th}}$
		  slice of the volume to segment, we consider three images,
		  corresponding to the $(n-1)^{\text{th}}$, $n^{\text{th}}$,
		  and $(n+1)^{\text{th}}$ slices of the original volume.
		  These three gray-level 2D images are assembled to form a 2D
		  RGB color image (one image per channel). This image is the
		  input of the FCN to obtain a 2D segmentation of the
		  $n^{\text{th}}$ slice. We process all slices, then stack
		  the results to form the 3D output segmentation. With such a
		  technique, the segmentation of the left atrial cavity on a
		  3D volume takes only a few seconds. We obtain a Dice score
		  of 0.92 both on the training set in our experiments before
		  the challenge, and on the test set of the challenge.}
}