Fig 1 - uploaded by Isuru Suranga Wijesinghe
Content may be subject to copyright.
Severity stages of Diabetic Retinopathy
Source publication
The automatic classification of diabetic retinopathy (DR) is of vital importance, as it is the leading cause of irreversible vision loss in the working-age population all over the world today. Current clinical approaches require a well-trained clinician to manually evaluate digital colour fundus photographs of retina and locate lesions associated w...
Context in source publication
Context 1
... to damage and loss of cells known as pericytes within the retina. DR will lead to blindness if untreated where as timely treatment can slow down or stop further vision loss. Therefore, people with diabetes should undergo regular eye screening for DR. Ophthalmologists and welltrained practitioners often use a five-class grading system as shown in Fig. 1 to describe the severity stages of DR, namely diabetes without retinopathy (Non-DR), Mild nonproliferative DR (Mild-NPDR), Moderate non-proliferative DR (Moderate-NPDR), Severe non-proliferative DR (Severe-NPDR) and Proliferative diabetic retinopathy (PDR) ...
Similar publications
Android is the most widely used mobile platform, making it a prime target for malicious attacks. Therefore, it is
imperative to effectively circumvent these attacks. Recently, machine learning has been a promising solution for
malware detection, which relies on distinguishing features. While machine learning-based malware scanners have
a large numb...
Citations
... After that, two types of classifiers, named deep convolutional neural network (DCNN) and deep full connection network (DFNN), were used for the final classification with an accuracy of 95.42%, which is also a result of texture-based information processing. Moreover, Wijesinghe et al. [10] proposed a transfer learning-based ensemble model consisting of DenseNet-201, ResNet-18, and VGG-16. The background removal, resolution optimization, and resizing were performed as image preprocessing to make the dataset more optimized for training. ...
The current advancement towards retinal disease detection mainly focused on distinct feature extraction using either a convolutional neural network (CNN) or a transformer-based end-to-end deep learning (DL) model. The individual end-to-end DL models are capable of only processing texture or shape-based information for performing detection tasks. However, extraction of only texture- or shape-based features does not provide the model robustness needed to classify different types of retinal diseases. Therefore, concerning these two features, this paper developed a fusion model called ‘Conv-ViT’ to detect retinal diseases from foveal cut optical coherence tomography (OCT) images. The transfer learning-based CNN models, such as Inception-V3 and ResNet-50, are utilized to process texture information by calculating the correlation of the nearby pixel. Additionally, the vision transformer model is fused to process shape-based features by determining the correlation between long-distance pixels. The hybridization of these three models results in shape-based texture feature learning during the classification of retinal diseases into its four classes, including choroidal neovascularization (CNV), diabetic macular edema (DME), DRUSEN, and NORMAL. The weighted average classification accuracy, precision, recall, and F1 score of the model are found to be approximately 94%. The results indicate that the fusion of both texture and shape features assisted the proposed Conv-ViT model to outperform the state-of-the-art retinal disease classification models.
... Deep neural networks work better on large datasets, and the size of the data set is a very important parameter in the network's performance. For this reason, most studies have used more than one dataset to improve classification [30,32,34,36,40,43,46,49,50,52,53,55,56,63,65,67]. Table 4 presents the number of classes for DR classification used in the reviewed papers. ...
Cases of diabetes and related diabetic retinopathy (DR) have been increasing at an alarming rate in modern times. Early detection of DR is an important problem since it may cause permanent blindness in the late stages. In the last two decades, many different approaches have been applied in DR detection. Reviewing academic literature shows that deep neural networks (DNNs) have become the most preferred approach for DR detection. Among these DNN approaches, Convolutional Neural Network (CNN) models are the most used ones in the field of medical image classification. Designing a new CNN architecture is a tedious and time-consuming approach. Additionally, training an enormous number of parameters is also a difficult task. Due to this reason, instead of training CNNs from scratch, using pre-trained models has been suggested in recent years as transfer learning approach. Accordingly, the present study as a review focuses on DNN and Transfer Learning based applications of DR detection considering 43 publications between 2015 and 2021. The published papers are summarized using 3 figures and 10 tables, giving information about 29 pre-trained CNN models, 13 DR data sets and standard performance metrics.
... Using pre-trained network without finetuning [30], [23] Fine-tuning entire pre-trained network [31], [32], [33], [34], [35], [36], [37], [38], [39], [22], [21], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [24], [53], [54], [55], [63] Fine-tuning a part of the pre-trained network [56], [34], [35] Training a state-of-art architecture from scratch [57], [34], [37], [40] Modifying a pre-trained network [58], [38], [42], [43], [45], [49], [55] Not stated [59], [60], [61] ...
... Using pre-trained network without finetuning [30], [23] Fine-tuning entire pre-trained network [31], [32], [33], [34], [35], [36], [37], [38], [39], [22], [21], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [24], [53], [54], [55], [63] Fine-tuning a part of the pre-trained network [56], [34], [35] Training a state-of-art architecture from scratch [57], [34], [37], [40] Modifying a pre-trained network [58], [38], [42], [43], [45], [49], [55] Not stated [59], [60], [61] ...
... Deep neural networks work better on large datasets, and the size of the data set is a very important parameter in the network's performance. For this reason, most studies have used more than one dataset to improve classification performance [35,36,43,45,51,61,63,59,31,33,39,53,41]. [36], [22], [21], [50], [51], [56], [37], [39], [58], [33], [31], [43], [49], [54], [63], [61], [64], [46], [59], [41] Non-public Dataset 10 [23], [44], [45], [47], [54], [57], [55], [38], [63], [52] Messidor Dataset 8 [36], [35], [63], [31], [60], [55], [39], [ Table 4 presents the number of classes for DR classification used in the reviewed papers. ...
Cases of diabetes and related diabetic retinopathy (DR) have been increasing at an alarming rate in modern times. Early detection of DR is an important problem since it may cause permanent blindness in the late stages. In the last two decades, many different approaches have been applied in DR detection. Reviewing academic literature shows that deep neural networks (DNNs) have become the most preferred approach for DR detection. Among these DNN approaches, Convolutional Neural Network (CNN) models are the most used ones in the field of medical image classification. Designing a new CNN architecture is a tedious and time-consuming approach. Additionally, training an enormous number of parameters is also a difficult task. Due to this reason, instead of training CNNs from scratch, using pre-trained models has been suggested in recent years as transfer learning approach. Accordingly, the present study as a review focuses on DNN and Transfer Learning based applications of DR detection considering 38 publications between 2015 and 2020. The published papers are summarized using 9 figures and 10 tables, giving information about 22 pre-trained CNN models, 12 DR data sets and standard performance metrics.
The current advancement towards retinal disease detection mainly focused on distinct
feature extraction using either a convolutional neural network (CNN) or a transformerbased
end-to-end deep learning (DL) model. The individual end-to-end DL models are
capable of only processing texture or shape-based information for performing detection
tasks. Thereby, concerning these two features, in this research, a fusion model is
developed which is called ‘Conv-ViT’ to detect retinal disease from foveal cut optical
coherence tomography (OCT) images. The transfer learning-based CNN models such as
Inception v3 and ResNet-50 are utilized to process texture information by calculating
the correlation of the nearby pixel. Additionally, the vision transformer model is fused
to process shape-based features by determining the correlation between long-distance
pixels. The hybridization of these three models results in shape-based texture feature
learning during the classification of retinal diseases into its four classes including
choroidal neovascularization (CNV), diabetic macular edema (DME), Drusen, and
Normal. The weighted average classification accuracy, precision, recall, and F1 score of
the model are found approximately 94%. The results indicate that the fusion of both
texture and shape features assisted the proposed Conv-ViT model to outperform the
state-of-the-art retinal disease classification models.
