GAN-Based Novel Approach for Data Augmentation with Improved Disease Classification

Deep learning, via Convolutional Neural Network (CNN) models, has had significant breakthroughs and achievements in image classification tasks where there is a sufficient amount of annotated data. Generally, medical image datasets are highly imbalanced, and training a convolutional neural network model to classify diseases across classes does not give the desired performance. To combat this, data augmentation is required, and in this chapter, we propose a new strategy to improve the generalization performance of CNN model using a novel online data augmentation strategy with Deep Convolutional Generative Adversarial Network (DCGAN). This helps in regularizing the training and gives better performance across classes, as it prevents the model from overfitting to the majority class. We performed our experiment on NIH chest X-ray image dataset, available openly, considering three classes: Infiltration, Atelectasis and No Findings. The test accuracy of the CNN model is 60.3% compared to the 65.3% test accuracy of the online GAN-augmented CNN model.

[1]  Jun-Geol Baek,et al.  Oversampling method using outlier detectable generative adversarial network , 2019, Expert Syst. Appl..

[2]  Soumith Chintala,et al.  Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks , 2015, ICLR.

[3]  Frans Coenen,et al.  Convolutional Neural Networks for Diabetic Retinopathy , 2016, MIUA.

[4]  Yang Tao,et al.  Early Detection of Tomato Spotted Wilt Virus by Hyperspectral Imaging and Outlier Removal Auxiliary Classifier Generative Adversarial Nets (OR-AC-GAN) , 2019, Scientific Reports.

[5]  Heung-Il Suk,et al.  Deep Learning in Medical Image Analysis. , 2017, Annual review of biomedical engineering.

[6]  Lars M. Mescheder,et al.  On the convergence properties of GAN training , 2018, ArXiv.

[7]  Fang Deng,et al.  Data augmentation in fault diagnosis based on the Wasserstein generative adversarial network with gradient penalty , 2020, Neurocomputing.

[8]  Eyad Elyan,et al.  MFC-GAN: Class-imbalanced dataset classification using Multiple Fake Class Generative Adversarial Network , 2019, Neurocomputing.

[9]  Qingming Huang,et al.  Semantic invariant cross-domain image generation with generative adversarial networks , 2018, Neurocomputing.

[10]  Simon Osindero,et al.  Conditional Generative Adversarial Nets , 2014, ArXiv.

[11]  Anmol Sharma,et al.  A CADe System for Gliomas in Brain MRI using Convolutional Neural Networks , 2018, ArXiv.

[12]  Léon Bottou,et al.  Wasserstein GAN , 2017, ArXiv.

[13]  Kai Ma,et al.  X2CT-GAN: Reconstructing CT From Biplanar X-Rays With Generative Adversarial Networks , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

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

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

[16]  Shahrokh Valaee,et al.  Generalization of Deep Neural Networks for Chest Pathology Classification in X-Rays Using Generative Adversarial Networks , 2017, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[17]  Youyi Zheng,et al.  Hair-GAN: Recovering 3D hair structure from a single image using generative adversarial networks , 2019, Vis. Informatics.

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

[19]  Hayit Greenspan,et al.  Cross-Modality Synthesis from CT to PET using FCN and GAN Networks for Improved Automated Lesion Detection , 2018, Eng. Appl. Artif. Intell..

[20]  Ronald M. Summers,et al.  ChestX-ray: Hospital-Scale Chest X-ray Database and Benchmarks on Weakly Supervised Classification and Localization of Common Thorax Diseases , 2019, Deep Learning and Convolutional Neural Networks for Medical Imaging and Clinical Informatics.

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

[22]  Tara N. Sainath,et al.  Improving deep neural networks for LVCSR using rectified linear units and dropout , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[23]  Jingrui He,et al.  Misc-GAN: A Multi-scale Generative Model for Graphs , 2019, Front. Big Data.

[24]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[25]  Martin Wattenberg,et al.  GAN Lab: Understanding Complex Deep Generative Models using Interactive Visual Experimentation , 2018, IEEE Transactions on Visualization and Computer Graphics.

[26]  Haibo He,et al.  A Novel Generative Model With Bounded-GAN for Reliability Classification of Gear Safety , 2019, IEEE Transactions on Industrial Electronics.

[27]  Hayit Greenspan,et al.  GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classification , 2018, Neurocomputing.

[28]  Qingmin Liao,et al.  FV-GAN: Finger Vein Representation Using Generative Adversarial Networks , 2019, IEEE Transactions on Information Forensics and Security.