A Deep Learning Interpretable Classifier for Diabetic Retinopathy Disease Grading

Abstract In this paper we present a diabetic retinopathy deep learning interpretable classifier. On one hand, it classifies retina images into different levels of severity with good performance. On the other hand, this classifier is able of explaining the classification results by assigning a score for each point in the hidden and input spaces. These scores indicate the pixel contribution to the final classification. To obtain these scores, we propose a new pixel-wise score propagation model that for every neuron, divides the observed output score into two components. With this method, the generated visual maps can be easily interpreted by an ophthalmologist in order to find the underlying statistical regularities that help to the diagnosis of this eye disease.

[1]  Xiaohui Xie,et al.  DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification , 2018, 2018 IEEE Winter Conference on Applications of Computer Vision (WACV).

[2]  Aïda Valls,et al.  Weighted kappa loss function for multi-class classification of ordinal data in deep learning , 2018, Pattern Recognit. Lett..

[3]  Matthew D. Davis,et al.  Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. , 2003, Ophthalmology.

[4]  Yoshua. Bengio,et al.  Learning Deep Architectures for AI , 2007, Found. Trends Mach. Learn..

[5]  Alexander Binder,et al.  On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation , 2015, PloS one.

[6]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[7]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

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

[9]  Ming Yang,et al.  Classification of Alzheimer’s Disease Based on Eight-Layer Convolutional Neural Network with Leaky Rectified Linear Unit and Max Pooling , 2018, Journal of Medical Systems.

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

[11]  Nadav Cohen,et al.  On the Expressive Power of Deep Learning: A Tensor Analysis , 2015, COLT 2016.

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

[13]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[14]  Aïda Valls,et al.  Diabetic Retinopathy Detection Through Image Analysis Using Deep Convolutional Neural Networks , 2016, CCIA.

[15]  Jacob Cohen,et al.  Weighted kappa: Nominal scale agreement provision for scaled disagreement or partial credit. , 1968 .

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

[17]  Sergey Ioffe,et al.  Rethinking the Inception Architecture for Computer Vision , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[18]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[19]  Subhashini Venugopalan,et al.  Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs. , 2016, JAMA.

[20]  Sebastian Thrun,et al.  Dermatologist-level classification of skin cancer with deep neural networks , 2017, Nature.

[21]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[22]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

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

[24]  Motoaki Kawanabe,et al.  How to Explain Individual Classification Decisions , 2009, J. Mach. Learn. Res..

[25]  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.

[26]  Ohad Shamir,et al.  The Power of Depth for Feedforward Neural Networks , 2015, COLT.

[27]  Sergey Ioffe,et al.  Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning , 2016, AAAI.

[28]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[29]  Raquel Urtasun,et al.  Understanding the Effective Receptive Field in Deep Convolutional Neural Networks , 2016, NIPS.

[30]  Xiang Zhang,et al.  OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks , 2013, ICLR.

[31]  Alexander Binder,et al.  Explaining nonlinear classification decisions with deep Taylor decomposition , 2015, Pattern Recognit..

[32]  Luca Maria Gambardella,et al.  Deep, Big, Simple Neural Nets for Handwritten Digit Recognition , 2010, Neural Computation.

[33]  Guy Cazuguel,et al.  FEEDBACK ON A PUBLICLY DISTRIBUTED IMAGE DATABASE: THE MESSIDOR DATABASE , 2014 .

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

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