Unravelling Diabetic Retinopathy Through Image Processing, Neural Networks and Fuzzy Logic – A Review

Srinivasan, A.; Sudha, S.

doi:10.22159/ajpcr.2017.v10i4.17023

Cited by 6 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Automatic screening systems based on neural networks have been used to distinguish between no referable DR and referable DR, and the latter would be referred to ophthalmologists for further examination and treatment (9,10). These automatic systems have good performance and could substantially reduce doctor's workload (11). However, there is no quantitative synthesis of diagnostic accuracy of these methods.…”

Section: Introductionmentioning

confidence: 99%

Performance of deep neural network-based artificial intelligence method in diabetic retinopathy screening: a systematic review and meta-analysis of diagnostic test accuracy

Wang

Zhang

Lei

et al. 2020

European Journal of Endocrinology

View full text Add to dashboard Cite

Objective Automatic diabetic retinopathy screening system based on neural networks has been used to detect diabetic retinopathy (DR). However, there is no quantitative synthesis of performance of these methods. We aimed to estimate the sensitivity and specificity of neural networks in DR grading. Methods Medline, Embase, IEEE Xplore, and Cochrane Library were searched up to 23 July 2019. Studies that evaluated performance of neural networks in detection of moderate or worse DR or diabetic macular edema using retinal fundus images with ophthalmologists’ judgment as reference standard were included. Two reviewers extracted data independently. Risk of bias of eligible studies was assessed using QUDAS-2 tool. Results Twenty-four studies involving 235 235 subjects were included. Quantitative random-effects meta-analysis using the Rutter and Gatsonis hierarchical summary receiver operating characteristics (HSROC) model revealed a pooled sensitivity of 91.9% (95% CI: 89.6% to 94.3%) and specificity of 91.3% (95% CI: 89.0% to 93.5%). Subgroup analyses and meta-regression did not provide any statistically significant findings for the heterogeneous diagnostic accuracy in studies with different image resolutions, sample sizes of training sets, architecture of convolutional neural networks, or diagnostic criteria. Conclusions State-of-the-art neural networks could effectively detect clinical significant DR. To further improve diagnostic accuracy of neural networks, researchers might need to develop new algorithms rather than simply enlarge sample sizes of training sets or optimize image quality.

show abstract

Section: Introductionmentioning

confidence: 99%

Performance of deep neural network-based artificial intelligence method in diabetic retinopathy screening: a systematic review and meta-analysis of diagnostic test accuracy

Wang

Zhang

Lei

et al. 2020

European Journal of Endocrinology

View full text Add to dashboard Cite

show abstract

“…7 These automatic systems have good performances and could substantially reduce doctor's workload. 8…”

Section: Introductionmentioning

confidence: 99%

“…7 These automatic systems have good performances and could substantially reduce doctor's workload. 8 We aimed to perform this systematic review and meta-analysis to quantitatively analyze the performance of the deep learning neural network in detecting DME in patients from OCT images and investigate the factors that affect the diagnostic accuracy.…”

Section: Introductionmentioning

confidence: 99%

Deep learning algorithms for detection of diabetic macular edema in OCT images: A systematic review and meta-analysis

Wang

Dong

et al. 2022

European Journal of Ophthalmology

View full text Add to dashboard Cite

Purpose Artificial intelligence (AI) can detect diabetic macular edema (DME) from optical coherence tomography (OCT) images. We aimed to evaluate the performance of deep learning neural networks in DME detection. Methods Embase, Pubmed, the Cochrane Library, and IEEE Xplore were searched up to August 14, 2021. We included studies using deep learning algorithms to detect DME from OCT images. Two reviewers extracted the data independently, and the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool was applied to assess the risk of bias. The study is reported according to Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies (PRISMA-DTA). Results Ninteen studies involving 41005 subjects were included. The pooled sensitivity and specificity were 96.0% (95% confidence interval (CI): 93.9% to 97.3%) and 99.3% (95% CI: 98.2% to 99.7%), respectively. Subgroup analyses found that data set selection, sample size of training set and the choice of OCT devices contributed to the heterogeneity (all P < 0.05). While there was no association between the diagnostic accuracy and transfer learning adoption or image management (all P > 0.05). Conclusions Deep learning methods, particularly the convolutional neural networks (CNNs) could effectively detect clinically significant DME, which can provide referral suggestions to the patients.

show abstract

Segmentation of Diabetic Retinopathy Based on Retinal Fundus Images Using Thresholding Technique

Ali¹,

Hamzah²,

Saad³

et al. 2022

Proceedings of the Multimedia University Engineering Conference (MECON 2022)

View full text Add to dashboard Cite

Diabetic Retinopathy is damage to the retina caused by complications of diabetes mellitus. Exudates are the primary sign of Diabetic Retinopathy identified on the opthalmoscope as white or yellowish areas with varying sizes, shapes and locations in the retina. Early detection can potentially reduce the risk of blindness. Diabetic retinopathy diagnosis for early-stage detection used a manual or semi-automated system that only an ophthalmologist can do it. This project proposed an automatic analysis of diabetic retinopathy and lesion of the eye in the retinal region using image processing techniques. The proposed technique is using thresholding technique. The data is taken from the DIRECTDB1 database. Preprocessing is applied first and then the image is segmented into 8 x 8 regions. An image histogram is then been calculated at each region to look for the maximum number of pixels for each intensity level. By comparing normal and exudate regions, the best threshold is found. The proposed technique is validated based on the accuracy, specificity, and sensitivity of exudate detection from the ground truth image reference. The mean from the range of the abnormal region of interest (ROI) is also calculated whether it is within range of the optimal value. In conclusion, adaptive thresholding is the method of choice to develop an automated detection of diabetic retinopathy. The result shows an accuracy of 65.79%, sensitivity is 31.58%, and specificity is 100%. The best segmentation and classification performance is achieved in the range of abnormal region ROI which is 0.35 to 0.55.

show abstract

Unravelling Diabetic Retinopathy Through Image Processing, Neural Networks and Fuzzy Logic – A Review

Cited by 6 publications

References 28 publications

Performance of deep neural network-based artificial intelligence method in diabetic retinopathy screening: a systematic review and meta-analysis of diagnostic test accuracy

Performance of deep neural network-based artificial intelligence method in diabetic retinopathy screening: a systematic review and meta-analysis of diagnostic test accuracy

Deep learning algorithms for detection of diabetic macular edema in OCT images: A systematic review and meta-analysis

Segmentation of Diabetic Retinopathy Based on Retinal Fundus Images Using Thresholding Technique

Contact Info

Product

Resources

About