Risk Identification of Diabetic Macular Edema Using E-Adoption of Emerging Technology

Risk Identification of Diabetic Macular Edema Using E-Adoption of Emerging Technology

Amit Kumar, Anand Shanker Tewari
Copyright: © 2022 |Pages: 20
DOI: 10.4018/IJEA.310000
OnDemand:
(Individual Articles)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

The accumulation of the blood leaks on the retina is known as diabetic macular edema (DME), which can result in irreversible blindness. Early diagnosis and therapy can stop DME. This study presents an e-adoption of emerging technology such as RadioDense model for detecting and classifying DME from retinal fundus images. The proposed model employs a modified version of DenseNet121, radiomics features, and the gradient boosting classifier. The authors evaluated many classifiers on the concatenated features. The efficacy of the classifier is determined by comparing each classifier's accuracy values. According to the evaluation results, the concatenated features extraction using gradient boosting classifier outperforms all other classifiers on the IDRiD dataset. For multi-class classification, the suggested electronic adoption of emerging technology such as RadioDense model outperformed these classifiers and attained an accuracy of 87.4%. It can help to decrease the strain of ophthalmologists diagnosing the DME during locking and unlocking the worldwide lockdown.
Article Preview
Top

Introduction

Diabetes is a prevalent disease that mainly affects the world's working-age population and is caused by elevated blood glucose levels (Reichel & Salz, 2015). The number of diabetic persons increased from 108 million in 1980 to 422 million in 2014. The world-health-organization (WHO) predicts that diabetes will be the seventh most prevalent incurable disease in the future (Krizhevsky et al., 2017). It is approximated that by 2030, the number of diabetic patients between the ages of 20 and 79 in India will increase to roughly 101.2 million (Whiting et al., 2011).

Diabetes causes retinal microvascular alterations that result in vision-threatening diseases. It causes fluid to leak onto the retina, and the retinal fluid seepage creates hard exudates. Diabetic macular edema (DME) refers to the retina with lesions such as hard exudates, which may cause irreversible vision loss (Giancardo et al., 2011; Hee et al., 1995). It is identified by an ocular examination of retinal fundus pictures for the availability of hard exudates, as shown in Figure 1.

Figure 1.

Diseased retinal image with aberrant anatomical structure

IJEA.310000.f01
Table 1.
DME categorization
DME caseFinding
case 0Hard Exudates lesions are not found (i.e., the retina is healthy)
case 1Hard Exudates lesions are detected outside the macular region (i.e., the retina is not healthy, and patients having such lesions at this particular location enter the primary stages of developing DME risk)
case 2Hard Exudates are detected within the macular region (i.e., the retina is not healthy, and patients having such lesions at this particular location entered the severe stage of developing DME risk)

The degree of DME is determined by the abnormalities, such as hard exudates, and their distance from the fovea or central macular region (Decencière et al., 2014). The cases of DME are detailed in Table 1 and illustrated in Figure 2. These images are selected from the IDRiD database (Porwal et al., 2018).

At least once every year, diabetes people must undergo a retinal checkup (Ferris, 1993; Kollias & Ulbig, 2010; Ting et al., 2016). The diagnosis of patients having diabetic eyes depends on the availability of ophthalmologists (Jones & Edwards, 2010; Lin et al., 2016). In metropolitan regions of India, one ophthalmologist is accessible per 9000 people; in rural areas, this ratio is approximately 608,000 (Raman et al., 2016). It is forecasted to have roughly 151 million diabetic patients, of which about 51 million will have diabetic retinopathy (DR) by 2045 in India (Atlas, 2015).

Complete Article List

Search this Journal:
Reset
Volume 16: 1 Issue (2024)
Volume 15: 2 Issues (2023)
Volume 14: 3 Issues (2022): 2 Released, 1 Forthcoming
Volume 13: 2 Issues (2021)
Volume 12: 2 Issues (2020)
Volume 11: 2 Issues (2019)
Volume 10: 2 Issues (2018)
Volume 9: 2 Issues (2017)
Volume 8: 2 Issues (2016)
Volume 7: 2 Issues (2015)
Volume 6: 2 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing