ABSTRACT
Brain extraction is an essential processing step for most brain magnetic resonance imaging (MRI) studies. Due to the inaccuracy of available labels in training dataset, existing methods based on U-Net can only obtain rough brains. In this paper, we propose a new deep-learning-based method for accurate 3D brain extraction, in which a U-Net model is trained in two stages using different loss functions. In the first stage, the binary cross entropy (BCE) loss is used to train the model with original head MRIs and coarse labelled brain masks as usual U-Net models. In the second stage, a composite loss function that integrates active contour model (ACM) and BCE loss is introduced to guide the further training. By this means, the final trained model can not only strip head scalp and skull from head MRI scans, but also remove cerebrospinal fluid around brain tissues. Both quantitative and qualitative test results show that our brain extraction is more accurate than other counterparts. The improvement enables to build better brain model with more details.
- M. A. Balafar, A. R. Ramli, M. I. Saripan, and S. Mashohor. 2010. Review of brain MRI image segmentation methods. Artificial Intelligence Review 33, 3 (March 2010), 261–274. https://doi.org/10.1007/s10462-010-9155-0.Google ScholarDigital Library
- Mark Jenkinson, Peter Bannister, Michael Brady, and Stephen Smith. 2002. Improved Optimization for the Robust and Accurate Linear Registration and Motion Correction of Brain Images. NeuroImage 17, 2 (October 2002), 825–841. https://doi.org/10.1006/nimg.2002.1132.Google ScholarCross Ref
- Simon S Keller and Neil Roberts. 2009. Measurement of brain volume using MRI: software, techniques, choices and prerequisites. Journal of anthropological sciences = Rivista di antropologia: JASS 87 (January 2009), 127-151. http://europepmc.org/abstract/MED/19663172.Google Scholar
- Anam Fatima, Ahmad Raza Shahid, Basit Raza, Tahir Mustafa Madni, and Uzair Iqbal Janjua. 2020. State-of-the-Art Traditional to the Machine and Deep-Learning-Based Skull Stripping Techniques, Models, and Algorithms. Digit Imaging 33, 6 (December 2020), 1443-1464. https://doi.org/10.1007/s10278-020-00367-5.Google Scholar
- David W. Shattuck, Stephanie R. Sandor-Leahy, Kirt A. Schaper, David A. Rottenberg, and Richard M. Leahy. 2001. Magnetic resonance image tissue classification using a partial volume model. Neuroimage 13, 5 (May 2001), 856-876. https://doi.org/10.1006/nimg.2000.0730.Google ScholarCross Ref
- Stephen M. Smith. Fast robust automated brain extraction. 2002. Human Brain Mapping 17, 3 (November 2002), 143-155. https://doi.org/10.1002/hbm.10062.Google ScholarCross Ref
- Juan Eugenio Iglesias, Cheng-Yi Liu, Paul M. Thompson, and Zhuowen Tu. 2011. Robust Brain Extraction Across Datasets and Comparison With Publicly Available Methods. IEEE transactions on medical imaging 30, 9 (April 2011), 1617-1634. https://doi.org/10.1109/TMI.2011.2138152Google ScholarCross Ref
- Jens Sjölund, Andreas Eriksson Järlideni, Mats Andersson, Hans Knutsson, and Håkan Nordström. 2014. Skull Segmentation in MRI by a Support Vector Machine Combining Local and Global Features. In Proceedings of the 2014 22nd International Conference on Pattern Recognition (ICPR '14). IEEE Computer Society, USA, 3274–3279. https://doi.org/10.1109/ICPR.2014.564.Google ScholarDigital Library
- Jens Kleesiek, Gregor Urban, Alexander Hubert, Daniel Schwarz, Klaus Maier-Hein, Martin Bendszus, and Armin Biller. 2016. Deep MRI brain extraction: A 3D convolutional neural network for skull stripping. NeuroImage 129 (April 2016), 460-469. https://doi.org/10.1016/j.neuroimage.2016.01.024.Google Scholar
- Hyunho Hwang, Hafiz Zia Ur Rehman, and Sungon Lee. 2019. 3D U-Net for Skull Stripping in Brain MRI. Applied Sciences 9, 3 (February 2019), 569. https://doi.org/10.3390/app9030569.Google ScholarCross Ref
- Oeslle Lucena, Roberto Souza, Letícia Rittner, Richard Frayne, and Roberto Lotufo. 2019. Convolutional neural networks for skull-stripping in brain MR imaging using Consensus-based Silver standard Masks. Artificial Intelligence in Medicine 98 (July 2019), 48-58. https://doi.org/10.1016/j.artmed.2019.06.008.Google Scholar
- Qian Zhang , Li Wang, Xiaopeng Zong, Weili Lin, Gang Li, and Dinggang Shen. 2019. Frnet: Flattened residual network for infant MRI skull stripping. 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019). 999-1002. https://doi.org/10.1109/ISBI.2019.8759167Google ScholarCross Ref
- Benjamin Puccio, James P. Pooley, John S. Pellman, Elise C. Taverna, and R. Cameron Craddock. 2016. The preprocessed connectomes project repository of manually corrected skull-stripped T1-weighted anatomical MRI data. Gigascience 5, 1 (October 2016), 45. https://doi.org/10.1186/s13742-016-0150-5.Google ScholarCross Ref
- Olaf Ronneberger, Philipp Fischer and Thomas Brox. 2015. U-Net: Convolutional Networks for Biomedical Image Segmentation. In Proceedings of the 2015 18th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCA ‘15). Spriger, Cham, Switzerland, 234-241. https://doi.org/10.1007/978-3-319-24574-4_28.Google ScholarCross Ref
- Kening Le, Zeyu Lou, and Xiaolin Tian. 2021. Nested Recurrent Residual Unet (NRRU) on GAN (NRRG) for Cardiac CT Images Segmentation Task. In 2021 2nd International Conference on Artificial Intelligence and Information Systems (ICAIIS 2021). Association for Computing Machinery, New York, NY, USA, Article 80, 1–5. https://doi.org/10.1145/3469213.3470281.Google Scholar
- Wenxuan Wang, Chen Chen, Meng Ding, Hong Yu, Sen Zha, and Jiangyun Li. 2021. TransBTS: Multimodal Brain Tumor Segmentation Using Transformer. Medical Image Computing and Computer Assisted Intervention – MICCAI 2021, 109-119. https://doi.org/10.1007/978-3-030-87193-2_11.Google Scholar
- Sheng Lu, Jungang Han, Jiantao Li, Liyang Zhu, Jiewei Jiang, and Shaojie Tang. 2021. Three-dimensional Medical Image Segmentation with SE-VNet Neural Networks. In 2021 3rd International Conference on Intelligent Medicine and Image Processing (IMIP ‘21). Association for Computing Machinery, New York, NY, USA, 14–20. https://doi.org/10.1145/3468945.3468948.Google Scholar
- Michael Kass, Andrew Witkin and Demetri Terzopoulos. 1988. Active contour models. Int. J. Comput. Vision 1, 4 (January 1988), 321-331. https://doi.org/10.1007/BF00133570.Google ScholarCross Ref
- Tony F. Chan, and Luminita A. Vese. 2001. Active contours without edges. IEEE Trans Image Process 10, 2 (February 2001), 266-77. https://doi.org/10.1109/83.902291.Google ScholarDigital Library
- Xu Chen, Bryan M. Williams, Srinivasa R. Vallabhaneni, Gabriela Czanner, Rachel Williams, and Yalin Zheng. 2019. Learning Active Contour Models for Medical Image Segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 11632-11640. https://doi.org/10.1109/CVPR.2019.01190.Google ScholarCross Ref
- Ali Hatamizadeh, Assaf Hoogi, Debleena Sengupta, Wuyue Lu, Brian Wilcox, Daniel Rubin, and Demetri Terzopoulos. 2019. Deep active lesion segmentation. International Workshop on Machine Learning in Medical Imaging, 98-105. https://doi.org/10.1007/978-3-030-32692-0_12.Google ScholarDigital Library
- Daniel P. Huttenlocher, Gregory A. Klanderman, and William J. Rucklidge. 1993. Comparing Images Using the Hausdorff Distance. IEEE Trans. Pattern Anal. Mach. Intell. 15, 9 (September 1993), 850–863. https://doi.org/10.1109/34.232073.Google ScholarDigital Library
- Juan Eugenio Iglesias. 2018. ROBEX 1.2. https://www.nitrc.org/projects/robex.Google Scholar
- Ron Kikinis, Steve D. Pieper, and Kirby G. Vosburgh. 2014. 3D Slicer: a platform for subject-specific image analysis, visualization, and clinical support. Intraoperative Imaging Image-Guided Therapy. https://www.slicer.org/.Google Scholar
- Accurate brain extraction on MRI using U-Net trained in two stages
Recommendations
Skull-Stripping of Glioblastoma MRI Scans Using 3D Deep Learning
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain InjuriesAbstractSkull-stripping is an essential pre-processing step in computational neuro-imaging directly impacting subsequent analyses. Existing skull-stripping methods have primarily targeted non-pathologically-affected brains. Accordingly, they may perform ...
An Adaptive 3D U-Net for White Matter, Gray Matter and Cerebrospinal Fluid Segmentation from 3D-Brain MRI
ICMLSC '21: Proceedings of the 2021 5th International Conference on Machine Learning and Soft ComputingMany methods for Alzheimer's disease detection in brain magnetic resonance imaging (MRI) is related to white matter (WM), grey matter (GM), and cerebrospinal fluid (CSF) regions. Therefore, in many neurological applications, the segmentation of these ...
Multi-resolution 3D CNN for MRI Brain Tumor Segmentation and Survival Prediction
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain InjuriesAbstractIn this study, an automated three dimensional (3D) deep segmentation approach for detecting gliomas in 3D pre-operative MRI scans is proposed. Then, a classification algorithm based on random forests, for survival prediction is presented. The ...
Comments