Brain Tumor Segmentation and Radiomics Survival Prediction: Contribution to the BRATS 2017 Challenge

By Fabian Isensee, P. Kickingereder, Wolfgang Wick, Martin Bendszus, Klaus Hermann Maier-Hein. BrainLes@MICCAI, 2017

538 citations recorded by the source corpus.

Quantitative analysis of brain tumors is critical for clinical decision making. While manual segmentation is tedious, time consuming and subjective, this task is at the same time very challenging to solve for automatic segmentation methods. In this paper we present our most recent effort on developing a robust segmentation algorithm in the form of a convolutional neural network. Our network architecture was inspired by the popular U-Net and has been carefully modified to maximize brain tumor segmentation performance. We use a dice loss function to cope with class imbalances and use extensive data augmentation to successfully prevent overfitting. Our method beats the current state of the art on BraTS 2015, is one of the leading methods on the BraTS 2017 validation set (dice scores of 0.896, 0.797 and 0.732 for whole tumor, tumor core and enhancing tumor, respectively) and achieves very good Dice scores on the test set (0.858 for whole, 0.775 for core and 0.647 for enhancing tumor). We furthermore take part in the survival prediction subchallenge by training an ensemble of a random forest regressor and multilayer perceptrons on shape features describing the tumor subregions. Our approach achieves 52.6% accuracy, a Spearman correlation coefficient of 0.496 and a mean square error of 209607 on the test set.

Subfields: Computer Vision and Pattern Recognition, Neurology, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/da74769ff7af1bb80ff89d730377a5328e1d7b71

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

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Deep learning based brain tumor segmentation: a survey (2020, methodology): “[51] reported a modified U-Net for brain tumor segmentation, where the authors used a dice loss function and extensive data augmentation to successfully prevent over-fitting.”
A review: Deep learning for medical image segmentation using multi-modality fusion (2019, methodology): “Most of the existing multi-modal medical image segmentation networks adopt the input-level fusion strategy, which directly integrates the multi-modal images in the original input space [54] [29] [30] [13] [71] [36] [77] [51] [12].”
No New-Net (2018, methodology): “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].”
Medical Image Segmentation Review: The Success of U-Net (2022, methodology): “Kickingereder et al. [93] studied brain tumor growth by a variant of 3D U-Net [94] compared with the conventional clinical procedure applied to MRI scans.”
Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge (2018, background): “6, 18057 Rostock, Germany 19 Tata Memorial Centre, Homi Bhabha National Institute, Mumbai, India 20 Shri Guru Gobind Singhji Institute of Engineering and Technology, Nanded, India 21 NVIDIA, Santa Clara, USA 22 Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, Ge”
3D Deep Learning on Medical Images: A Review (2020, background): “Isensee et al. [49] Proposed a 3D U-Net architecture which consists of a perspective collection pathway for brain tumor segmentation.”
Multimodal Brain Tumor Classification Using Deep Learning and Robust Feature Selection: A Machine Learning Application for Radiologists (2020, background): “Sample images collected from Multimodal Brain Dataset (BRATS) [15].”
Automated brain extraction of multisequence MRI using artificial neural networks (2019, result): “put brain mask was linearly upsampled to the original resolution for evaluation. Network Architecture The network architecture (depicted in Figure 1) shares similarities with our recent contribution (Isensee, et al., 2017) to the BraTS 2017 challenge (Menze, et al., 2015). It is inspired by the succ”
Generalizing Deep Learning for Medical Image Segmentation to Unseen Domains via Deep Stacked Transformation (2020, methodology): “Gamma correction is used in a highly competitive brain MRI segmentation algorithm [30], and contributes to the robust segmentation performance across multiple hospitals [31].”
Autofocus Layer for Semantic Segmentation (2018, methodology): “Our model is only surpassed by the pipelines of [10] and [7], who both used ensembles of CNNs with deep supervision and more aggressive data augmentation.”
TransMed: Transformers Advance Multi-Modal Medical Image Classification (2021, methodology): “[8]–[10] mainly used the input-level fusion, which is the most common fusion method in multi-modal medical image analysis.”