DenseHyper: an automatic recognition system for detection of hypertensive retinopathy using dense features transform and deep-residual learning

High blood pressure and diabetes are associated with a retinal abnormality known as Hypertensive Retinopathy (HR). The severity-level and duration of hypertension are straightly related to the incidence of HR-eye disease. The HR damages the pathological lesions of eyes such as arteriolar narrowing, retinal hemorrhage, macular edema, cotton wool spots, and blood vessels. In the early stages, it is important to detect and diagnose HR to prevent eye blindness. Currently, there are few computerize systems developed to recognize HR. However, those systems focused on extracting features through hand-craft and deep-learning models (DLMs) based techniques. As a result, the complex image processing algorithms are required in case of hand-crafted features and it is difficult to define generalized features by DLMs to recognize HR. Moreover, the classification accuracy is not up-to-the-mark even though by using deep-feature techniques as observed in state-of-the-art HR diagnostics systems. To solve these problems, a novel hypertensive retinopathy (DenseHyper) system is developed to detect the HR based on a proposed trained features layer (TF-L) and dense feature transform layer (DFT-L) to the deep residual learning (DRL) methods. The DenseHyper system consists of different multilayer dense architecture by integrating of TF-L by convolutional neural network (CNN) to learn features from different lesions, and generate specialized features by DFT-L. To develop DenseHyper system, a learning based dense feature transform (DFT) approach was integrated to increase classification accuracy. Three online sources besides one private data are gathered to test and compare the DenseHyper system. To show the performance of the DenseHyper system, the statistical analysis is also performed on 4270 retinal fundus images through sensitivity (SE), specificity (SP), accuracy (ACC) and area under the receiver operating curve (AUC) metrics. The significant results were achieved compare to state-of-the-art methods. On average, the SE of 93%, SP of 95%, ACC of 95% and 0.96 of AUC values were obtained through a 10-fold cross-validation test. Experimental results confirm the applicability of the DenseHyper system to accurately diagnosis of hypertensive retinopathy.

[1]  Sven Loncaric,et al.  Detection of exudates in fundus photographs using convolutional neural networks , 2015, 2015 9th International Symposium on Image and Signal Processing and Analysis (ISPA).

[2]  Suchismita Goswami,et al.  Automatic Measurement and Analysis of Vessel Width in Retinal Fundus Image , 2017 .

[3]  Carla Agurto,et al.  Detection of hypertensive retinopathy using vessel measurements and textural features , 2014, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[4]  Jinkai Cui,et al.  Retinal vessel segmentation in colour fundus images using Extreme Learning Machine , 2017, Comput. Medical Imaging Graph..

[5]  Lu Wang,et al.  Retinal Image Enhancement Using Robust Inverse Diffusion Equation and Self-Similarity Filtering , 2016, PloS one.

[6]  Zenghui Wang,et al.  Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review , 2017, Neural Computation.

[7]  Muhammad Moazam Fraz,et al.  Arterioles and Venules Classification in Retinal Images Using Fully Convolutional Deep Neural Network , 2018, ICIAR.

[8]  M. Usman Akram,et al.  Automated system for the detection of hypertensive retinopathy , 2014, 2014 4th International Conference on Image Processing Theory, Tools and Applications (IPTA).

[9]  Qaisar Abbas,et al.  Automatic recognition of severity level for diagnosis of diabetic retinopathy using deep visual features , 2017, Medical & Biological Engineering & Computing.

[10]  Dahlan Abdullah,et al.  Hypertensive retinopathy identification through retinal fundus image using backpropagation neural network , 2018 .

[11]  Mark D. Huffman,et al.  Executive Summary: Heart Disease and Stroke Statistics—2015 Update A Report From the American Heart Association , 2011, Circulation.

[12]  Max A. Viergever,et al.  Ridge-based vessel segmentation in color images of the retina , 2004, IEEE Transactions on Medical Imaging.

[13]  Greg Russell,et al.  DR HAGIS—a fundus image database for the automatic extraction of retinal surface vessels from diabetic patients , 2017, Journal of medical imaging.

[14]  Anjan Gudigar,et al.  Deep convolution neural network for accurate diagnosis of glaucoma using digital fundus images , 2018, Inf. Sci..

[15]  Erik J. Bekkers,et al.  Automatic detection of vascular bifurcations and crossings in retinal images using orientation scores , 2016, 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI).

[16]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[17]  Aldenor G. Santos,et al.  Occurrence of the potent mutagens 2- nitrobenzanthrone and 3-nitrobenzanthrone in fine airborne particles , 2019, Scientific Reports.

[18]  Hong Qin,et al.  Video Saliency Detection via Spatial-Temporal Fusion and Low-Rank Coherency Diffusion , 2017, IEEE Transactions on Image Processing.

[19]  Julian Quiroga,et al.  Support system for the preventive diagnosis of Hypertensive Retinopathy , 2010, 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology.

[20]  Yan Liu,et al.  Deep residual learning for image steganalysis , 2018, Multimedia Tools and Applications.

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

[22]  M. Usman Akram,et al.  Classification of retinal vessels into arteries and veins for detection of hypertensive retinopathy , 2014, 2014 Cairo International Biomedical Engineering Conference (CIBEC).

[23]  Himansu Das,et al.  Performance Evaluation of Different Machine Learning Methods and Deep-Learning Based Convolutional Neural Network for Health Decision Making , 2019 .

[24]  M. Usman Akram,et al.  Decision support system for detection of hypertensive retinopathy using arteriovenous ratio , 2018, Artif. Intell. Medicine.

[25]  Alfredo Ruggeri,et al.  Computer estimation of the AVR parameter in diabetic retinopathy , 2009 .

[26]  Manoranjan Paul,et al.  Deep Learning Models for Retinal Blood Vessels Segmentation: A Review , 2019, IEEE Access.

[27]  M. Usman Akram,et al.  Automated detection of Cotton Wool Spots for the diagnosis of Hypertensive Retinopathy , 2014, 2014 Cairo International Biomedical Engineering Conference (CIBEC).

[28]  Christopher M O'Connor,et al.  Treatment of Hypertension in Patients With Coronary Artery Disease: A Scientific Statement from the American Heart Association, American College of Cardiology, and American Society of Hypertension. , 2015, Journal of the American College of Cardiology.

[29]  Nikolaos Doulamis,et al.  Deep Learning for Computer Vision: A Brief Review , 2018, Comput. Intell. Neurosci..

[30]  Weihong Deng,et al.  Very deep convolutional neural network based image classification using small training sample size , 2015, 2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR).

[31]  Kevin Noronha,et al.  Support System for the Automated Detection of Hypertensive Retinopathy using Fundus Images , 2013 .

[32]  Jiang Liu,et al.  Automatic localization of optic disc based on deep learning in fundus images , 2017, 2017 IEEE 2nd International Conference on Signal and Image Processing (ICSIP).

[33]  Saeid Nahavandi,et al.  A Classifier Graph Based Recurring Concept Detection and Prediction Approach , 2018, Comput. Intell. Neurosci..

[34]  U. Rajendra Acharya,et al.  Segmentation of optic disc, fovea and retinal vasculature using a single convolutional neural network , 2017, J. Comput. Sci..

[35]  Joni-Kristian Kämäräinen,et al.  The DIARETDB1 Diabetic Retinopathy Database and Evaluation Protocol , 2007, BMVC.

[36]  Saeed Sharifian,et al.  Modified deep residual network architecture deployed on serverless framework of IoT platform based on human activity recognition application , 2019, Future Gener. Comput. Syst..

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

[38]  Alicja R. Rudnicka,et al.  Automated arteriole and venule classification using deep learning for retinal images from the UK Biobank cohort , 2017, Comput. Biol. Medicine.

[39]  Muhammad Hussain,et al.  Deep Learning based Computer-Aided Diagnosis Systems for Diabetic Retinopathy: A Survey , 2018, Artif. Intell. Medicine.

[40]  K. Abbaspour,et al.  The future of extreme climate in Iran , 2019, Scientific Reports.

[41]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[42]  Na Li,et al.  Learning-Based Visual Saliency Model for Detecting Diabetic Macular Edema in Retinal Image , 2016, Comput. Intell. Neurosci..

[43]  Sotirios A. Tsaftaris,et al.  2019 11th International Symposium on Image and Signal Processing and Analysis (ISPA) , 2013 .

[44]  M. Nath,et al.  Detection of changes in color fundus images due to diabetic retinopathy , 2012, 2012 2nd National Conference on Computational Intelligence and Signal Processing (CISP).

[45]  Indrin J Chetty,et al.  Automatic Segmentation of the Prostate on CT Images Using Deep Neural Networks (DNN). , 2019, International journal of radiation oncology, biology, physics.

[46]  Myeongsu Kang,et al.  Deep Residual Networks With Dynamically Weighted Wavelet Coefficients for Fault Diagnosis of Planetary Gearboxes , 2018, IEEE Transactions on Industrial Electronics.

[47]  Manuel G. Penedo,et al.  Development of an automated system to classify retinal vessels into arteries and veins , 2012, Comput. Methods Programs Biomed..

[48]  Lixin Zheng,et al.  Combining Convolutional Neural Network With Recursive Neural Network for Blood Cell Image Classification , 2018, IEEE Access.

[49]  Chong Peng,et al.  Improved Saliency Detection in RGB-D Images Using Two-Phase Depth Estimation and Selective Deep Fusion , 2020, IEEE Transactions on Image Processing.

[50]  Qaisar Abbas,et al.  DermoDeep-A classification of melanoma-nevus skin lesions using multi-feature fusion of visual features and deep neural network , 2019, Multimedia Tools and Applications.

[51]  Enrico Grisan,et al.  A Novel Method for the Automatic Grading of Retinal Vessel Tortuosity , 2008, IEEE Transactions on Medical Imaging.

[52]  K. Vijayarekha,et al.  Hypertensive Retinopathy Diagnosis from Fundus Images by Estimation of Avr. , 2012 .

[53]  Hiroshi Fujita,et al.  Automated selection of major arteries and veins for measurement of arteriolar-to-venular diameter ratio on retinal fundus images , 2011, Comput. Medical Imaging Graph..

[54]  Chengdong Wu,et al.  Automatic Optic Disc Segmentation Based on Modified Local Image Fitting Model with Shape Prior Information , 2019, Journal of healthcare engineering.

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

[56]  Vasudevan Lakshminarayanan,et al.  Ophthalmic diagnosis using deep learning with fundus images - A critical review , 2020, Artif. Intell. Medicine.

[57]  Eugenio Culurciello,et al.  An Analysis of Deep Neural Network Models for Practical Applications , 2016, ArXiv.

[58]  M. Usman Akram,et al.  Arteriovenous ratio and papilledema based hybrid decision support system for detection and grading of hypertensive retinopathy , 2018, Comput. Methods Programs Biomed..

[59]  Wiharto,et al.  The review of computer aided diagnostic hypertensive retinopathy based on the retinal image processing , 2019, IOP Conference Series: Materials Science and Engineering.

[60]  Francesco Calimeri,et al.  Novel Method for Automated Analysis of Retinal Images: Results in Subjects with Hypertensive Retinopathy and CADASIL , 2015, BioMed research international.

[61]  Widodo Budiharto,et al.  The Classification of Hypertensive Retinopathy using Convolutional Neural Network , 2017, ICCSCI.

[62]  Li-Qun Xu,et al.  Convolutional Neural Network for Retinal Blood Vessel Segmentation , 2016, 2016 9th International Symposium on Computational Intelligence and Design (ISCID).

[63]  J. Recio-Rodríguez,et al.  Plasma Cardiotrophin-1 as a Marker of Hypertension and Diabetes-Induced Target Organ Damage and Cardiovascular Risk , 2015, Medicine.

[64]  Qaisar Abbas,et al.  A comprehensive review of recent advances on deep vision systems , 2018, Artificial Intelligence Review.

[65]  Kostas Marias,et al.  An image analysis framework for the early assessment of hypertensive retinopathy signs , 2011, 2011 E-Health and Bioengineering Conference (EHB).

[66]  Qaisar Abbas,et al.  Video scene analysis: an overview and challenges on deep learning algorithms , 2017, Multimedia Tools and Applications.

[67]  Oscar Camacho Nieto,et al.  A machine learning approach to medical image classification: Detecting age-related macular degeneration in fundus images , 2017, Comput. Electr. Eng..

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

[69]  D. Squirrell,et al.  Detection of smoking status from retinal images; a Convolutional Neural Network study , 2019, Scientific Reports.

[70]  Y D Pradipto,et al.  Detection of Hypertension Retinopathy Using Deep Learning and Boltzmann Machines , 2017 .

[71]  Hiroshi Ito,et al.  Modern treatment to reduce pulmonary arterial pressure in pulmonary arterial hypertension. , 2018, Journal of cardiology.

[72]  J. Wainer,et al.  Advancing Bag-of-Visual-Words Representations for Lesion Classification in Retinal Images , 2014, PloS one.