Exudate detection in fundus images using deeply-learnable features
| Autor: | Behzad Aliahmad, Dinesh Kumar, Tiago Carvalho, Leandro Aparecido Passos Junior, Edmar Rezende, João Paulo Papa, Parham Khojasteh |
|---|---|
| Přispěvatelé: | School of Engineering, Universidade Federal de São Carlos (UFSCar), Federal Institute of São Paulo, Universidade Estadual de Campinas (UNICAMP), Universidade Estadual Paulista (Unesp) |
| Rok vydání: | 2018 |
| Předmět: |
0301 basic medicine
Exudate Databases Factual Computer science Fundus Oculi Boltzmann machine Health Informatics Fundus (eye) Convolutional neural network 03 medical and health sciences 0302 clinical medicine Deep Learning Discriminative model Diabetic retinopathy medicine Image Processing Computer-Assisted Humans Tomography Optical Sensitivity (control systems) Exudate detection Deep residual networks Discriminative restricted Boltzmann machines Diabetic Retinopathy business.industry Deep learning Pattern recognition Computer Science Applications Support vector machine 030104 developmental biology Convolutional neural networks Artificial intelligence medicine.symptom business 030217 neurology & neurosurgery |
| Zdroj: | Scopus Repositório Institucional da UNESP Universidade Estadual Paulista (UNESP) instacron:UNESP |
| ISSN: | 1879-0534 |
| Popis: | Made available in DSpace on 2019-10-06T16:54:37Z (GMT). No. of bitstreams: 0 Previous issue date: 2019-01-01 Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Presence of exudates on a retina is an early sign of diabetic retinopathy, and automatic detection of these can improve the diagnosis of the disease. Convolutional Neural Networks (CNNs) have been used for automatic exudate detection, but with poor performance. This study has investigated different deep learning techniques to maximize the sensitivity and specificity. We have compared multiple deep learning methods, and both supervised and unsupervised classifiers for improving the performance of automatic exudate detection, i.e., CNNs, pre-trained Residual Networks (ResNet-50) and Discriminative Restricted Boltzmann Machines. The experiments were conducted on two publicly available databases: (i) DIARETDB1 and (ii) e-Ophtha. The results show that ResNet-50 with Support Vector Machines outperformed other networks with an accuracy and sensitivity of 98% and 0.99, respectively. This shows that ResNet-50 can be used for the analysis of the fundus images to detect exudates. Royal Melbourne Institute of Technology Biosignals Laboratory School of Engineering, 124 La Trobe St Federal University of São Carlos Department of Computing, Rod. Washington Luís, Km 235 Federal Institute of São Paulo Department of Computing University of Campinas Institute of Computing São Paulo State University - UNESP Department of Computing, Av. Eng. Luiz Edmundo Carrijo Coube, 14-01 São Paulo State University - UNESP Department of Computing, Av. Eng. Luiz Edmundo Carrijo Coube, 14-01 FAPESP: #2013/07375-0 FAPESP: #2014/12236-1 FAPESP: #2016/19403-6 FAPESP: #2016/50022-9 CNPq: #307066/2017-7 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect