2022
DOI: 10.3390/sym14071427
|View full text |Cite
|
Sign up to set email alerts
|

AI-Based Automatic Detection and Classification of Diabetic Retinopathy Using U-Net and Deep Learning

Abstract: Artificial intelligence is widely applied to automate Diabetic retinopathy diagnosis. Diabetes-related retinal vascular disease is one of the world’s most common leading causes of blindness and vision impairment. Therefore, automated DR detection systems would greatly benefit the early screening and treatment of DR and prevent vision loss caused by it. Researchers have proposed several systems to detect abnormalities in retinal images in the past few years. However, Diabetic Retinopathy automatic detection met… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
25
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
6
3
1

Relationship

0
10

Authors

Journals

citations
Cited by 94 publications
(25 citation statements)
references
References 52 publications
0
25
0
Order By: Relevance
“…Finally, the IoTDL-DRD approach has categorized fundus image into severe classes with accu y , sens y , spec y , F score , AUC score , and MCC of 99.43%, 89.47%, 100%, 94.44%, 94.74%, and 94.31% correspondingly. For assuring the enhanced DR classification performance of the IoTDL-DRD model, a wide ranging comparative analysis is made in Table V and Figure 13 [25]. The fallouts highlighted the IoTDL-DRD model has shown enhanced outcomes over other models [26].…”
Section: Resultsmentioning
confidence: 99%
“…Finally, the IoTDL-DRD approach has categorized fundus image into severe classes with accu y , sens y , spec y , F score , AUC score , and MCC of 99.43%, 89.47%, 100%, 94.44%, 94.74%, and 94.31% correspondingly. For assuring the enhanced DR classification performance of the IoTDL-DRD model, a wide ranging comparative analysis is made in Table V and Figure 13 [25]. The fallouts highlighted the IoTDL-DRD model has shown enhanced outcomes over other models [26].…”
Section: Resultsmentioning
confidence: 99%
“…Deep neural networks can learn rich information about visual features of classes that appear in images when trained on vast amounts of labeled data. These attributes significantly contributed to various critical applications, including medical applications 1 , 2 .However, their ability to generalize to new classes diminishes when presented with only a limited number of labeled examples 3 , which is a prevalent issue in domains such as geospatial and medical, where collecting and labeling large datasets is a complex and expensive process. To overcome this issue, researchers have proposed the few-shot learning paradigm, which attempts to mimic the capacity of the human visual system to rapidly learn new classes from a small number of labeled examples.…”
Section: Introductionmentioning
confidence: 99%
“…It is important for individuals with diabetes to undergo regular eye exams to detect any early signs of DR and receive appropriate treatment [6]. Despite its increasing prevalence globally, there remains a need for automated diagnostic tools to enhance the early detection and prevention of DR [7]. Various methods have been employed to study the highly precise system for the diagnosis of DR, including feature extraction using deep neural networks, retinal image classification, and lesion detection and segmentation.…”
Section: Introductionmentioning
confidence: 99%