Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 208.46 KB
Add to list
  1. Search results
  2. Fully automatic binary glioma grading...

Fully automatic binary glioma grading based on pre-therapy MRI using 3D Convolutional Neural Networks

Milan Decuyper (UGent) and Roel Van Holen (UGent)
(2019) Medical Imaging with Deep Learning, MIDL 2019, Proceedings.
Author
Organization
Abstract
The optimal treatment strategy of newly diagnosed glioma is strongly influenced by tumour malignancy. Manual non-invasive grading based on MRI is not always accurate and biopsies to verify diagnosis negatively impact overall survival. In this paper, we propose a fully automatic 3D computer-aided diagnosis (CAD) system to non-invasively differentiate high-grade glioblastoma from lower-grade glioma. The approach consists of an automatic segmentation step to extract the tumour ROI followed by classification using a 3D convolutional neural network. Segmentation was performed using a 3D U-Net achieving a dice score of 88.53% which matches top performing algorithms in the BraTS 2018 challenge. The classification network was trained and evaluated on a large heterogeneous dataset of 549 patients reaching an accuracy of 91%. Additionally, the CAD system was evaluated on data from the Ghent University Hospital and achieved an accuracy of 92% which shows that the algorithm is robust to data from different centres.
Keywords
Deep Learning, CNN, Glioma Grading, MRI

Downloads

  • Download
    1908.01506v1.pdf
    • full text (Published version)
    • |
    • open access
    • |
    • PDF
    • |
    • 208.46 KB

Citation

Please use this url to cite or link to this publication:

MLA
Decuyper, Milan, and Roel Van Holen. “Fully Automatic Binary Glioma Grading Based on Pre-Therapy MRI Using 3D Convolutional Neural Networks.” Medical Imaging with Deep Learning, MIDL 2019, Proceedings, edited by Jorge Cardoso et al., 2019.
APA
Decuyper, M., & Van Holen, R. (2019). Fully automatic binary glioma grading based on pre-therapy MRI using 3D Convolutional Neural Networks. In J. Cardoso, A. Feragen, B. Glocker, E. Konukoglu, I. Oguz, G. Unal, & T. Vercauteren (Eds.), Medical Imaging with Deep Learning, MIDL 2019, Proceedings. London, UK.
Chicago author-date
Decuyper, Milan, and Roel Van Holen. 2019. “Fully Automatic Binary Glioma Grading Based on Pre-Therapy MRI Using 3D Convolutional Neural Networks.” In Medical Imaging with Deep Learning, MIDL 2019, Proceedings, edited by Jorge Cardoso, Aasa Feragen, Ben Glocker, Ender Konukoglu, Ipek Oguz, Gozde Unal, and Tom Vercauteren. London, UK.
Chicago author-date (all authors)
Decuyper, Milan, and Roel Van Holen. 2019. “Fully Automatic Binary Glioma Grading Based on Pre-Therapy MRI Using 3D Convolutional Neural Networks.” In Medical Imaging with Deep Learning, MIDL 2019, Proceedings, ed by. Jorge Cardoso, Aasa Feragen, Ben Glocker, Ender Konukoglu, Ipek Oguz, Gozde Unal, and Tom Vercauteren. London, UK.
Vancouver
1.
Decuyper M, Van Holen R. Fully automatic binary glioma grading based on pre-therapy MRI using 3D Convolutional Neural Networks. In: Cardoso J, Feragen A, Glocker B, Konukoglu E, Oguz I, Unal G, et al., editors. Medical Imaging with Deep Learning, MIDL 2019, Proceedings. London, UK; 2019.
IEEE
[1]
M. Decuyper and R. Van Holen, “Fully automatic binary glioma grading based on pre-therapy MRI using 3D Convolutional Neural Networks,” in Medical Imaging with Deep Learning, MIDL 2019, Proceedings, London, UK, 2019.
@inproceedings{8627042,
  abstract     = {{The optimal treatment strategy of newly diagnosed glioma is strongly influenced by tumour malignancy. Manual non-invasive grading based on MRI is not always accurate and biopsies to verify diagnosis negatively impact overall survival. In this paper, we propose a fully automatic 3D computer-aided diagnosis (CAD) system to non-invasively differentiate high-grade glioblastoma from lower-grade glioma. The approach consists of an automatic segmentation step to extract the tumour ROI followed by classification using a 3D convolutional neural network. Segmentation was performed using a 3D U-Net achieving a dice score of 88.53% which matches top performing algorithms in the BraTS 2018 challenge. The classification network was trained and evaluated on a large heterogeneous dataset of 549 patients reaching an accuracy of 91%. Additionally, the CAD system was evaluated on data from the Ghent University Hospital and achieved an accuracy of 92% which shows that the algorithm is robust to data from different centres.}},
  author       = {{Decuyper, Milan and Van Holen, Roel}},
  booktitle    = {{Medical Imaging with Deep Learning, MIDL 2019, Proceedings}},
  editor       = {{Cardoso, Jorge and Feragen, Aasa and Glocker, Ben and Konukoglu, Ender and Oguz, Ipek and Unal, Gozde and Vercauteren, Tom}},
  keywords     = {{Deep Learning,CNN,Glioma Grading,MRI}},
  language     = {{eng}},
  location     = {{London, UK}},
  pages        = {{4}},
  title        = {{Fully automatic binary glioma grading based on pre-therapy MRI using 3D Convolutional Neural Networks}},
  url          = {{https://arxiv.org/html/1907.08612}},
  year         = {{2019}},
}