Automated Glioma Grading on Conventional MRI images Using Deep Convolutional Neural Networks.

PURPOSE Gliomas are the most common primary tumor of the brain and are classified into grades I-IV of the World Health Organization (WHO), based on their invasively histological appearance. Gliomas grading plays an important role to determine the treatment plan and prognosis prediction. In this study we propose two novel methods for automatic, non-invasively distinguishing low-grade (Grades II and III) glioma (LGG) and high-grade (grade IV) glioma (HGG) on conventional MRI images by using deep convolutional neural networks (CNNs). METHODS All MRI images have been preprocessed first by rigid image registration and intensity inhomogeneity correction. Both proposed methods consist of two steps: (1) 3D brain tumor segmentation based on a modification of the popular U-Net model; (2) tumor classification on segmented brain tumor. In the first method, the slice with largest area of tumor is determined and the state-of-the-art mask R-CNN model is employed for tumor grading. To improve the performance of the grading model, a 2D data augmentation has been implemented to increase both the amount and the diversity of the training images. In the second method, denoted as 3DConvNet, a 3D volumetric CNNs is applied directly on bounding image regions of segmented tumor for classification, which can fully leverage the 3D spatial contextual information of volumetric image data. RESULTS The proposed schemes were evaluated on The Cancer Imaging Archive (TCIA) low grade glioma (LGG) data, and the Multimodal Brain Tumor Image Segmentation (BraTS) Benchmark 2018 training data sets with 5-fold cross validation. All data are divided into training, validation, and test sets. Based on biopsy-proven ground truth, the performance metrics of sensitivity, specificity, and accuracy are measured on the test sets. The results are 0.935 (sensitivity), 0.972 (specificity), and 0.963 (accuracy) for the 2D Mask R-CNN based method, and 0.947 (sensitivity), 0.968 (specificity), and 0.971 (accuracy) for the 3DConvNet method, respectively. In regard to efficiency, for 3D brain tumor segmentation, the program takes around ten and a half hours for training with 300 epochs on BraTS 2018 data set and takes only around 50 seconds for testing of a typical image with a size of 160x216x176. For 2D Mask R-CNN based tumor grading, the program takes around 4 hours for training with around 60000 iterations, and around 1 second for testing of a 2D slice image with size of 128x128. For 3DConvNet based tumor grading, the program takes around 2 hours for training with 10000 iterations, and 0.25 seconds for testing of a 3D cropped image with size of 64x64x64, using a DELL PRECISION Tower T7910, with two NVIDIA Titan Xp GPUs. CONCLUSIONS Two effective glioma grading methods on conventional MRI images using deep convolutional neural networks have been developed. Our methods are fully automated without manual specification of region-of-interests and selection of slices for model training, which are common in traditional machine learning based brain tumor grading methods. This methodology may play a crucial role in selecting effective treatment options and survival predictions without the need for surgical biopsy.

[1]  S. Cha,et al.  Update on brain tumor imaging: from anatomy to physiology. , 2006, AJNR. American journal of neuroradiology.

[2]  Parashkev Nachev,et al.  Computer Methods and Programs in Biomedicine NiftyNet: a deep-learning platform for medical imaging , 2022 .

[3]  Geethu Mohan,et al.  MRI based medical image analysis: Survey on brain tumor grade classification , 2018, Biomed. Signal Process. Control..

[4]  Michael Brady,et al.  Estimating the bias field of MR images , 1997, IEEE Transactions on Medical Imaging.

[5]  Christos Davatzikos,et al.  Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme , 2009, Magnetic resonance in medicine.

[6]  Vinod Kumar,et al.  A dual neural network ensemble approach for multiclass brain tumor classification , 2012, International journal for numerical methods in biomedical engineering.

[7]  Subhashis Banerjee,et al.  Deep Radiomics for Brain Tumor Detection and Classification from Multi-Sequence MRI , 2019, ArXiv.

[8]  W El-Deredy,et al.  Tumour grading from magnetic resonance spectroscopy: a comparison of feature extraction with variable selection , 2003, Statistics in medicine.

[9]  Massimo Caulo,et al.  Data-driven grading of brain gliomas: a multiparametric MR imaging study. , 2014, Radiology.

[10]  Jayaram K. Udupa,et al.  Intensity standardization simplifies brain MR image segmentation , 2009, Comput. Vis. Image Underst..

[11]  Frank G Zoellner,et al.  Predictive modeling in glioma grading from MR perfusion images using support vector machines , 2008, Magnetic resonance in medicine.

[12]  Hao Chen,et al.  VoxResNet: Deep voxelwise residual networks for brain segmentation from 3D MR images , 2017, NeuroImage.

[13]  Kaiming He,et al.  Feature Pyramid Networks for Object Detection , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

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

[15]  C. Davatzikos,et al.  Survival Analysis of Patients with High-Grade Gliomas Based on Data Mining of Imaging Variables , 2012, American Journal of Neuroradiology.

[16]  Bowen Xin,et al.  Machine Learning Models for Multiparametric Glioma Grading With Quantitative Result Interpretations , 2019, Front. Neurosci..

[17]  Hidenao Fukuyama,et al.  Voxel-based clustered imaging by multiparameter diffusion tensor images for glioma grading , 2014, NeuroImage: Clinical.

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

[19]  Carles Arús,et al.  Automated classification of short echo time in in vivo 1H brain tumor spectra: A multicenter study , 2003, Magnetic resonance in medicine.

[20]  S. Bauer,et al.  A survey of MRI-based medical image analysis for brain tumor studies , 2013, Physics in medicine and biology.

[21]  B. Scheithauer,et al.  The 2007 WHO classification of tumours of the central nervous system , 2007, Acta Neuropathologica.

[22]  J. Barnholtz-Sloan,et al.  The epidemiology of glioma in adults: a "state of the science" review. , 2014, Neuro-oncology.

[23]  Jirí Sedlár,et al.  Predicting Deletion of Chromosomal Arms 1p/19q in Low-Grade Gliomas from MR Images Using Machine Intelligence , 2017, Journal of Digital Imaging.

[24]  Boqiang Liu,et al.  Computer-Aided Grading of Gliomas Combining Automatic Segmentation and Radiomics , 2018, Int. J. Biomed. Imaging.

[25]  Quoc V. Le,et al.  AutoAugment: Learning Augmentation Strategies From Data , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Brian B. Avants,et al.  N4ITK: Improved N3 Bias Correction , 2010, IEEE Transactions on Medical Imaging.

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

[28]  Yang Yang,et al.  Glioma Grading on Conventional MR Images: A Deep Learning Study With Transfer Learning , 2018, Front. Neurosci..

[29]  Guy Cosnard,et al.  Multiparametric magnetic resonance imaging to differentiate high-grade gliomas and brain metastases. , 2012, Journal of neuroradiology. Journal de neuroradiologie.

[30]  H. R. Arvinda,et al.  RETRACTED ARTICLE: Glioma grading: sensitivity, specificity, positive and negative predictive values of diffusion and perfusion imaging , 2009, Journal of Neuro-Oncology.

[31]  Jayaram K. Udupa,et al.  New variants of a method of MRI scale standardization , 2000, IEEE Transactions on Medical Imaging.

[32]  Li Fei-Fei,et al.  ImageNet: A large-scale hierarchical image database , 2009, CVPR.

[33]  Kaiming He,et al.  Mask R-CNN , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

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

[35]  Victor Alves,et al.  Brain Tumor Segmentation Using Convolutional Neural Networks in MRI Images , 2016, IEEE Transactions on Medical Imaging.

[36]  Gustavo Carneiro,et al.  A Bayesian Data Augmentation Approach for Learning Deep Models , 2017, NIPS.

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

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

[39]  G. Reifenberger,et al.  The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary , 2016, Acta Neuropathologica.

[40]  Sanjay Ranka,et al.  Visual Explanations From Deep 3D Convolutional Neural Networks for Alzheimer's Disease Classification , 2018, AMIA.

[41]  Klaus H. Maier-Hein,et al.  Brain Tumor Segmentation and Radiomics Survival Prediction: Contribution to the BRATS 2017 Challenge , 2017, BrainLes@MICCAI.