Automatic Brain Tumor Segmentation Using Cascaded Anisotropic Convolutional Neural Networks

By Guotai Wang, Wenqi Li, Sébastien Ourselin, Tom Kamiel Magda Vercauteren. 2017

546 citations recorded by the source corpus.

A cascade of fully convolutional neural networks is proposed to segment multi-modal Magnetic Resonance (MR) images with brain tumor into background and three hierarchical regions: whole tumor, tumor core and enhancing tumor core. The cascade is designed to decompose the multi-class segmentation problem into a sequence of three binary segmentation problems according to the subregion hierarchy. The whole tumor is segmented in the first step and the bounding box of the result is used for the tumor core segmentation in the second step. The enhancing tumor core is then segmented based on the bounding box of the tumor core segmentation result. Our networks consist of multiple layers of anisotropic and dilated convolution filters, and they are combined with multi-view fusion to reduce false positives. Residual connections and multi-scale predictions are employed in these networks to boost the segmentation performance. Experiments with BraTS 2017 validation set show that the proposed method achieved average Dice scores of 0.7859, 0.9050, 0.8378 for enhancing tumor core, whole tumor and tumor core, respectively. The corresponding values for BraTS 2017 testing set were 0.7831, 0.8739, and 0.7748, respectively.

Subfields: Computer Vision and Pattern Recognition, Neurology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Neurology

Citations in the corpus

Image Segmentation Using Deep Learning: A Survey (2020, methodology): “FCNs have been applied to a variety of segmentation problems, such as brain tumor segmentation [31], instanceaware semantic segmentation [32], skin lesion segmenta-”
Medical Image Synthesis for Data Augmentation and Anonymization using Generative Adversarial Networks (2018, methodology): “[20] (trained on real and synthetic data) are shown in Table 1.”
Interactive Medical Image Segmentation Using Deep Learning With Image-Specific Fine Tuning (2017, background): “Compared with large isotropic 3D receptive fields [8], it has less memory consumption [26].”
3D MRI brain tumor segmentation using autoencoder regularization (2018, unknown): “ training they used a batch size of 8, and a crop of 64x64x64 3D patch. EMMA’s ensemble of dierent models demonstrated a good generalization performance winning the BraTS 2017 challenge. Wang et al. [21] second place paper took a dierent approach, by training 3 networks for each tumor subregion in c”
Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge (2018, background): “S eg m en ta tio n 1 biomedia1 Konstantinos Kamnitsas Imperial College London, UK [33] 2 UCL-TIG Guotai Wang University College London (UCL), UK [55] 3 (tie) MIC DKFZ Fabian Isensee Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, Germany [29]”
Brain Tumor Detection Based on Deep Learning Approaches and Magnetic Resonance Imaging (2023, methodology): “In addition, multiclass segmentation of three tumor regions was performed using cascaded CNNs in [58].”
No New-Net (2018, background): “Recently, a growing number of architectural modiﬁcations to encoder-decoder networks have been proposed that are designed to improve the performance of the networks for their speciﬁc tasks [17,19,20,21,10,6,22,7].”
Deep learning based brain tumor segmentation: a survey (2020, methodology): “CNN Single-Path [26], [27], [28], [29], [30], [31], [32] Multi-Path [33], [34], [35], [36], [37], [38], [39], [40], [41], [42] FCN Vanilla FCN [43], [44], [28], [45], [46], [47], [48], [49], [50] U/V-Net Based [51], [52], [53], [54], [55], [56], [57], [58], [59], [60] Cascaded CNN [61], [62], [63], ”
3D Anisotropic Hybrid Network: Transferring Convolutional Features from 2D Images to 3D Anisotropic Volumes (2017, methodology): “Similar to our work, all the isotropic convolutions are replaced by 3 × 3 × 1 and 1 × 1 × 3 convolutions in [17, 21].”