DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis
Diabetic macular edema (DME) is the most common cause of irreversible vision loss in diabetes patients. Early diagnosis of DME is necessary for effective treatment of the disease. Visual detection of DME in retinal screening images by ophthalmologists is a time-consuming process. Recently, many comp...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-09-01
|
Series: | Entropy |
Subjects: | |
Online Access: | https://www.mdpi.com/1099-4300/23/10/1251 |
_version_ | 1797514630134759424 |
---|---|
author | Ghada Atteia Nagwan Abdel Samee Hassan Zohair Hassan |
author_facet | Ghada Atteia Nagwan Abdel Samee Hassan Zohair Hassan |
author_sort | Ghada Atteia |
collection | DOAJ |
description | Diabetic macular edema (DME) is the most common cause of irreversible vision loss in diabetes patients. Early diagnosis of DME is necessary for effective treatment of the disease. Visual detection of DME in retinal screening images by ophthalmologists is a time-consuming process. Recently, many computer-aided diagnosis systems have been developed to assist doctors by detecting DME automatically. In this paper, a new deep feature transfer-based stacked autoencoder neural network system is proposed for the automatic diagnosis of DME in fundus images. The proposed system integrates the power of pretrained convolutional neural networks as automatic feature extractors with the power of stacked autoencoders in feature selection and classification. Moreover, the system enables extracting a large set of features from a small input dataset using four standard pretrained deep networks: ResNet-50, SqueezeNet, Inception-v3, and GoogLeNet. The most informative features are then selected by a stacked autoencoder neural network. The stacked network is trained in a semi-supervised manner and is used for the classification of DME. It is found that the introduced system achieves a maximum classification accuracy of 96.8%, sensitivity of 97.5%, and specificity of 95.5%. The proposed system shows a superior performance over the original pretrained network classifiers and state-of-the-art findings. |
first_indexed | 2024-03-10T06:34:19Z |
format | Article |
id | doaj.art-075aca9328e2497ea6ee760b5cec704d |
institution | Directory Open Access Journal |
issn | 1099-4300 |
language | English |
last_indexed | 2024-03-10T06:34:19Z |
publishDate | 2021-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Entropy |
spelling | doaj.art-075aca9328e2497ea6ee760b5cec704d2023-11-22T18:10:06ZengMDPI AGEntropy1099-43002021-09-012310125110.3390/e23101251DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME DiagnosisGhada Atteia0Nagwan Abdel Samee1Hassan Zohair Hassan2Information Technology Department, College of Computer and Information Sciences, Princess Nourah Bint Abdulrahman University, Riyadh 11461, Saudi ArabiaInformation Technology Department, College of Computer and Information Sciences, Princess Nourah Bint Abdulrahman University, Riyadh 11461, Saudi ArabiaDepartment of Mechanical Engineering, College of Engineering, Alfaisal University, Takhassusi Street, P.O. Box 50927, Riyadh 11533, Saudi ArabiaDiabetic macular edema (DME) is the most common cause of irreversible vision loss in diabetes patients. Early diagnosis of DME is necessary for effective treatment of the disease. Visual detection of DME in retinal screening images by ophthalmologists is a time-consuming process. Recently, many computer-aided diagnosis systems have been developed to assist doctors by detecting DME automatically. In this paper, a new deep feature transfer-based stacked autoencoder neural network system is proposed for the automatic diagnosis of DME in fundus images. The proposed system integrates the power of pretrained convolutional neural networks as automatic feature extractors with the power of stacked autoencoders in feature selection and classification. Moreover, the system enables extracting a large set of features from a small input dataset using four standard pretrained deep networks: ResNet-50, SqueezeNet, Inception-v3, and GoogLeNet. The most informative features are then selected by a stacked autoencoder neural network. The stacked network is trained in a semi-supervised manner and is used for the classification of DME. It is found that the introduced system achieves a maximum classification accuracy of 96.8%, sensitivity of 97.5%, and specificity of 95.5%. The proposed system shows a superior performance over the original pretrained network classifiers and state-of-the-art findings.https://www.mdpi.com/1099-4300/23/10/1251diabetic macular edemaretinal fundus imagedeep learningpretrained convolutional neural networkautoencodertransfer learning |
spellingShingle | Ghada Atteia Nagwan Abdel Samee Hassan Zohair Hassan DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis Entropy diabetic macular edema retinal fundus image deep learning pretrained convolutional neural network autoencoder transfer learning |
title | DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis |
title_full | DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis |
title_fullStr | DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis |
title_full_unstemmed | DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis |
title_short | DFTSA-Net: Deep Feature Transfer-Based Stacked Autoencoder Network for DME Diagnosis |
title_sort | dftsa net deep feature transfer based stacked autoencoder network for dme diagnosis |
topic | diabetic macular edema retinal fundus image deep learning pretrained convolutional neural network autoencoder transfer learning |
url | https://www.mdpi.com/1099-4300/23/10/1251 |
work_keys_str_mv | AT ghadaatteia dftsanetdeepfeaturetransferbasedstackedautoencodernetworkfordmediagnosis AT nagwanabdelsamee dftsanetdeepfeaturetransferbasedstackedautoencodernetworkfordmediagnosis AT hassanzohairhassan dftsanetdeepfeaturetransferbasedstackedautoencodernetworkfordmediagnosis |