Fully Convolutional Deep Residual Neural Networks for Brain Tumor Segmentation

In this paper, a fully convolutional residual neural network (FCR-NN) based on linear identity mappings is implemented for medical image segmentation, employed here in the setting of brain tumors. Inspired by deep residual networks which won the ImageNet ILSVRC 2015 classification challenge, the FCR-NN combines optimization gains from residual identity mappings with a fully convolutional architecture for image segmentation that efficiently accounts for both low- and high-level image features. After training two separate networks, one for the task of whole tumor segmentation and a second for tissue sub-region segmentation, the serial FCR-NN architecture exceeds state-of-the art with complete tumor, core tumor and enhancing tumor validation Dice scores of 0.87, 0.81 and 0.72 respectively. Despite each FCR-NN comprising a complex 22 layer architecture, the fully convolutional design allows for complete segmentation of a tumor volume within 2 s.

[1]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[2]  Christopher Joseph Pal,et al.  Brain tumor segmentation with Deep Neural Networks , 2015, Medical Image Anal..

[3]  John Wong,et al.  Anti-PD-1 blockade and stereotactic radiation produce long-term survival in mice with intracranial gliomas. , 2013, International journal of radiation oncology, biology, physics.

[4]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[5]  Razvan Pascanu,et al.  Advances in optimizing recurrent networks , 2012, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[6]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[7]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[8]  Brian B. Avants,et al.  The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) , 2015, IEEE Transactions on Medical Imaging.

[9]  Danilo P. Mandic,et al.  A generalized normalized gradient descent algorithm , 2004, IEEE Signal Processing Letters.

[10]  Luca Maria Gambardella,et al.  Deep Neural Networks Segment Neuronal Membranes in Electron Microscopy Images , 2012, NIPS.

[11]  Victor Alves,et al.  Deep Convolutional Neural Networks for the Segmentation of Gliomas in Multi-sequence MRI , 2015, Brainles@MICCAI.

[12]  Bilwaj Gaonkar,et al.  GLISTRboost: Combining Multimodal MRI Segmentation, Registration, and Biophysical Tumor Growth Modeling with Gradient Boosting Machines for Glioma Segmentation , 2015, Brainles@MICCAI.

[13]  R. Meier,et al.  A Hybrid Model for Multimodal Brain Tumor Segmentation , 2013 .

[14]  Nicholas Ayache,et al.  A Generative Model for Brain Tumor Segmentation in Multi-Modal Images , 2010, MICCAI.

[15]  Stephen T. C. Wong,et al.  Semi-automatic segmentation software for quantitative clinical brain glioblastoma evaluation. , 2012, Academic radiology.

[16]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[17]  Gábor Székely,et al.  A Generative Probabilistic Model and Discriminative Extensions for Brain Lesion Segmentation— With Application to Tumor and Stroke , 2016, IEEE Transactions on Medical Imaging.

[18]  Hao Chen,et al.  VoxResNet: Deep Voxelwise Residual Networks for Volumetric Brain Segmentation , 2016, ArXiv.

[19]  Rasoul Khayati,et al.  Fully automated tumor segmentation based on improved fuzzy connectedness algorithm in brain MR images , 2011, Comput. Biol. Medicine.

[20]  Guido Gerig,et al.  Simulation of brain tumors in MR images for evaluation of segmentation efficacy , 2009, Medical Image Anal..

[21]  Vinod Kumar,et al.  A novel content-based active contour model for brain tumor segmentation. , 2012, Magnetic resonance imaging.

[22]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  Gözde B. Ünal,et al.  Tumor-Cut: Segmentation of Brain Tumors on Contrast Enhanced MR Images for Radiosurgery Applications , 2012, IEEE Transactions on Medical Imaging.

[24]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[25]  Andrea Vedaldi,et al.  MatConvNet: Convolutional Neural Networks for MATLAB , 2014, ACM Multimedia.

[26]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[27]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[28]  Yoshua Bengio,et al.  Convolutional networks for images, speech, and time series , 1998 .

[29]  Klaus H. Maier-Hein,et al.  Deep MRI brain extraction: A 3D convolutional neural network for skull stripping , 2016, NeuroImage.

[30]  Songfan Yang,et al.  Multi-scale Recognition with DAG-CNNs , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[31]  Koenraad Van Leemput,et al.  Brain Tumor Segmentation Using a Generative Model with an RBM Prior on Tumor Shape , 2015, Brainles@MICCAI.

[32]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).