GENTEL : GENerating Training data Efficiently
for Learning to segment medical images

Rajat Prince Thakur,Sergi Pujades Rocamora,Lavika Goel,Rolf Pohmann,Jürgen Machann,Michael. J. Black

Abstract

Accurately segmenting MRI images is crucial for many clinical applications. However, manually segmenting images with accurate pixel precision is a tedious and time consuming task. In this paper we present a simple, yet effective method to improve the efficiency of the image segmentation process. We propose to transform the image annotation task into a binary choice task. We start by using classical image processing algorithms with different parameter values to generate multiple, different segmentation masks for each input MRI image. Then, instead of segmenting the pixels of the images, the user only needs to decide whether a segmentation is acceptable or not. This method allows us to efficiently obtain high quality segmentations with minor human intervention. With the selected segmentations, we train a state-of-the-art neural network model. For the evaluation, we use a second MRI dataset (1.5T Dataset), acquired with a different protocol and containing annotations. We show that the trained network i) is able to automatically segment cases where none of the classical methods obtain a high quality result; ii) generalizes to the second MRI dataset, which was acquired with a different protocol and was never seen at training time ; and iii) enables detection of miss-annotations in this second dataset. Quantitatively, the trained network obtains very good results : DICE score - mean 0.98, median 0.99- and Hausdorff distance (in pixels) - mean 4.7, median 2.0-.

Paper PDF

RFIAP_2020_camera_ready.pdf

Citing the GENTEL paper

@conference{gentel:rfiap:2020,
  title = {{GENTEL : GENerating Training data Efficiently for Learning to segment medical images}},
  author = {Thakur, Rajat Prince and Rocamora, Sergi Pujades and Goel, Lavika and Pohmann, Rolf and Machann, J\"{u}rgen and Black, Michael J.},
  booktitle = {Congr\`{e}s Reconnaissance des Formes, Image, Apprentissage et Perception (RFAIP)},
  month = jun,
  year = {2020},
  month_numeric = {6}
}

Authors and affiliations

Rajat Prince Thakur¹,³ Sergi Pujades Rocamora² Lavika Goel⁶ Rolf Pohmann⁴ Jürgen Machann⁵ Michael J. Black¹

¹ Max Planck Institute for Intelligent Systems, Tübingen, Germany

² Université Grenoble Alpes, Inria, CNRS, Grenoble INP, LJK, France

³ Birla Institute for Technology and Science, Pilani, India

⁴ Max Planck Institute for Biological Cybernetics, Tübingen, Germany

⁵ Eberhard Karls University of Tübingen, Tübingen, Germany

⁶ Malaviya National Institute of Technology (NIT), Jaipur, India

Contact

For questions, please contact gentel@tue.mpg.de.