Learning High-Resolution and Efficient Non-local Features for Brain Glioma Segmentation in MR Images

Brain glioma segmentation using multi-parametric magnetic resonance (MR) imaging has significant clinical value. Although 3D convolutional neural networks (CNNs) have become increasingly prevalent in delivering this segmentation task, these models still suffer from an insufficient ability to high-resolution feature representation for small and irregular regions, limited local receptive fields, and poor long-range dependencies. In this paper, we propose a 3D High-resolution and Non-local Feature Network (HNF-Net) for brain glioma segmentation using multi-parametric MR imaging. We construct HNF-Net based mainly on the parallel multi-scale fusion (PMF) module, which helps produce strong high-resolution feature representation and aggregate multi-scale contextual information. We also introduce the expectation-maximization attention (EMA) module to HNF-Net, aiming to capture the long-range dependent contextual information and reduce the feature redundancy in a lightweight fashion. We evaluated our HNF-Net on the BraTS 2019 Challenge dataset against eight top-ranking methods listed on the challenge leaderboard. Our results suggest that the proposed HNF-Net achieves improved overall performance over these methods, and our ablation study demonstrates the effectiveness of the PMF module and EMA module.

[1]  Abhinav Gupta,et al.  Non-local Neural Networks , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[2]  Andriy Myronenko,et al.  3D MRI brain tumor segmentation using autoencoder regularization , 2018, BrainLes@MICCAI.

[3]  et al.,et al.  Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge , 2018, ArXiv.

[4]  Jakob Verbeek,et al.  Convolutional Neural Fabrics , 2016, NIPS.

[5]  Klaus H. Maier-Hein,et al.  No New-Net , 2018, 1809.10483.

[6]  Thomas Brox,et al.  3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation , 2016, MICCAI.

[7]  Yi Zhang,et al.  PSANet: Point-wise Spatial Attention Network for Scene Parsing , 2018, ECCV.

[8]  Konstantinos Kamnitsas,et al.  Efficient multi‐scale 3D CNN with fully connected CRF for accurate brain lesion segmentation , 2016, Medical Image Anal..

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

[10]  Bastian Leibe,et al.  Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Seyed-Ahmad Ahmadi,et al.  V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[12]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[13]  Nima Tajbakhsh,et al.  UNet++: A Nested U-Net Architecture for Medical Image Segmentation , 2018, DLMIA/ML-CDS@MICCAI.

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

[15]  Guang Yang,et al.  Automatic Brain Tumor Detection and Segmentation Using U-Net Based Fully Convolutional Networks , 2017, MIUA.

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

[17]  Sébastien Ourselin,et al.  Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations , 2017, DLMIA/ML-CDS@MICCAI.

[18]  Hong Liu,et al.  Expectation-Maximization Attention Networks for Semantic Segmentation , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[19]  Christos Davatzikos,et al.  Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features , 2017, Scientific Data.

[20]  Chen Chen,et al.  3D Dilated Multi-Fiber Network for Real-time Brain Tumor Segmentation in MRI , 2019, MICCAI.

[21]  Jun Fu,et al.  Dual Attention Network for Scene Segmentation , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).