A Novel Automated System for Early Diabetic Retinopathy Detection and Severity Classification DOI Open Access

Santoshkumar S Ainapur,

Vedashri Patil

Journal of Chemometrics, Journal Year: 2024, Volume and Issue: 38(11)

Published: Aug. 19, 2024

ABSTRACT Diabetes is a common and serious global disease that damages blood vessels in the eye, leading to vision loss. Early accurate diagnosis of this issue crucial reduce risk visual impairment. The typical deep learning (DL) methods for diabetic retinopathy (DR) grading are often time‐consuming, resulting unsatisfactory detection performance due inadequate representation lesion features. To overcome these challenges, research proposes new automated mechanism detecting classifying DR, aiming identify DR severities different stages. figure out capture feature characteristics from samples, conjugated attention transformer utilized within collective net model, which automatically generates maps diagnosing DR. These extracted then fused through fusion function calculating weights produce most powerful map. Finally, cases identified discriminated using kernel extreme machine (KELM) model. For evaluating severity, our work utilizes four benchmark datasets: APTOS 2019, MESSIDOR‐2 dataset, DiaRetDB1 V2.1, DIARETDB0 datasets. illuminate data noise unwanted variations, two preprocessing steps carried out, include contrast enhancement illumination correction. experimental results evaluated well‐known indicators demonstrate suggested method achieves higher accuracy 99.63% compared other baseline methods. This contributes development screening techniques less time‐consuming capable identifying severity levels at premature level.

Language: Английский

Diabetic Retinopathy Classification With Deep Learning via Fundus Images: A Short Survey DOI Creative Commons
Shanshan Zhu,

Changchun Xiong,

Qingshan Zhong

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 20540 - 20558

Published: Jan. 1, 2024

Diabetic retinopathy (DR) is a microvascular disease that associated with diabetes mellitus. DR can cause irreversible vision loss and blindness. classification, is, early diagnosis accurate grading, critical for protection immediate treatment. Deep learning-based automated systems led to significant expectations classification based on fundus images several advantages. In the past years, many outstanding studies in this area have been conducted review articles published. However, new trends future directions are need further analyzed. Thus, we carefully included read 94 related published from 2018 2023 through Web of Science, PubMed, Scopus, IEEE Xplore. From review, found transfer learning has used as an strategy overcoming issue limited data resources support analysis. CNN models ResNet VGGNet layers tens or even hundreds most popular frameworks classification. The APTOS 2019 EyePACS widely datasets addition, some lightweight DL architectures like SqueezeNet MobileNet proposed tasks, especially computational capabilities. Although deep achieved surpassed human-level accuracy there still long way go real clinical workflows. Further improvements model interpretability, trustworthiness ophthalmologists, cost-effective reliable screening needed.

Language: Английский

Citations

11
Diabetic retinopathy detection via exudates and hemorrhages segmentation using iterative NICK thresholding, watershed, and Chi2 feature ranking DOI Creative Commons

Patsaphon Chandhakanond,

Pakinee Aimmanee

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 14, 2025

Language: Английский

Citations

0
Reliable and Energy-Efficient Diabetic Retinopathy Screening Using Memristor-Based Neural Networks DOI Creative Commons
Sumit Diware,

Koteswararao Chilakala,

Rajiv Joshi

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 47469 - 47482

Published: Jan. 1, 2024

Diabetic retinopathy (DR) is a leading cause of permanent vision loss worldwide. It refers to irreversible retinal damage caused due elevated glucose levels and blood pressure. Regular screening for DR can facilitate its early detection timely treatment. Neural network-based classifiers be leveraged achieve such in convenient automated manner. However, these suffer from reliability issue where they exhibit strong performance during development but degraded after deployment. Moreover, do not provide supplementary information about the prediction outcome, which severely limits their widespread adoption. Furthermore, energy-efficient deployment on edge devices remains unaddressed, crucial enhance global accessibility. In this paper, we present reliable hardware detection, suitable devices. We first develop classification model using custom training data that incorporates diverse image quality sources along with improved class balance. This enables our effectively handle both on-field variations images minority classes, enhancing post-deployment reliability. then propose pseudo-binary scheme further improve prediction. Additionally, an design memristor-based computation-in-memory, Our proposed approach achieves three orders magnitude reduction energy consumption over state-of-the-art platforms.

Language: Английский

Citations

4
Automatic Diabetic Retinopathy Identification Using the Zernike Moment Decomposition in Fundus Images DOI
Javier Almeida,

I. J. Orlando Guerrero

IFMBE proceedings, Journal Year: 2025, Volume and Issue: unknown, P. 259 - 269

Published: Jan. 1, 2025

Language: Английский

Citations

0
Optimal Convolutional Networks for Staging and Detecting of Diabetic Retinopathy DOI Creative Commons
Minyar Sassi Hidri, Adel Hidri, Suleiman Ali Alsaif

et al.

Information, Journal Year: 2025, Volume and Issue: 16(3), P. 221 - 221

Published: March 13, 2025

Diabetic retinopathy (DR) is the main ocular complication of diabetes. Asymptomatic for a long time, it subject to annual screening using dilated fundus or retinal photography look early signs. Fundus and optical coherence tomography (OCT) are used by ophthalmologists assess thickness structure, as well detect edema, hemorrhage, scarring. The effectiveness ConvNet no longer needs be demonstrated, its use in field imaging has made possible overcome many barriers, which were until now insurmountable with old methods. Throughout this study, robust optimal deep proposed analyze images automatically distinguish between healthy, moderate, severe DR. model combines architecture taken from ImageNet, data augmentation, class balancing, transfer learning order establish benchmarking test. A significant improvement at level middle corresponds stage DR, was major problem previous studies. By eliminating need retina specialists broadening access care, substantially more objectively staging detecting

Language: Английский

Citations

0
Diabetic Retinopathy Detection Using DL‐Based Feature Extraction and a Hybrid Attention‐Based Stacking Ensemble DOI Creative Commons

Sanjana Rajeshwar,

Shreya Thaplyal,

M. S. Anbarasi

et al.

Advances in Public Health, Journal Year: 2025, Volume and Issue: 2025(1)

Published: Jan. 1, 2025

Diabetic retinopathy (DR) poses a significant threat to vision if left undetected and untreated. This paper addresses this challenge by utilizing advanced deep learning (DL) algorithms with established image processing techniques enhance accuracy efficiency in detection. Image extracts critical features from retinal images, acting as early warning signs for DR. Our proposed hybrid model combines machine (ML) strengths, leveraging discriminative abilities custom features. The methodology involves data acquisition diverse dataset, augmentation enrich training data, multistep pipeline. Feature extraction utilizes ResNet50, InceptionV3, visual geometry group (VGG)‐19 their outputs classification. Classification employs decision tree (DT), K‐nearest neighbor (KNN), support vector (SVM), modified convolutional neural network (CNN) spatial attention layer. work attention‐based stacking ensemble the mentioned models base layer logistic regression meta layer, which further enhanced accuracy. system, evaluated through metrics like confusion matrix, accuracy, receiver operating characteristic (ROC) curve, promises improved diagnostic capabilities. yields an of 99.768%.

Language: Английский

Citations

0
Predicting Diabetic Retinopathy Severity with Deep Learning: A Survey of Fundus Image Analysis Technique DOI

A Binusha Sornil,

C. Sheeja Herobin Rani,

I.Rexilin Sheeba

et al.

Published: April 12, 2024

Diabetic retinopathy (DR) is a significant complication of diabetes mellitus, impacting vision due to retinal abnormalities. Early detection and precise severity assessment are crucial for effective management. Leveraging deep learning techniques image preprocessing methods, this paper proposes comprehensive approach DR classification. Utilizing publicly available datasets like EyePACS, Messidor-2, APTOS, DDR, steps including Gaussian blurring data augmentation employed enhance quality address class imbalance. Wavelet decomposition used feature extraction capture multi-resolution information from fundus images. Transfer with ResNet variants, coupled regularization techniques, aids in model generalization. A modified ResNet50 architecture introduced, featuring custom fully connected layers additional convolutional improved extraction. The aims classify diseases into four levels: normal, mild, moderate, severe proliferative. survey aspect delves methods' effectiveness improving CNN performance medical analysis, specifically detection. applicability transfer imaging tasks, particularly DR, also explored. This study contributes advancing analysis diagnosis classification, addressing the critical need efficient management debilitating condition.

Language: Английский

Citations

0
Automated Diabetic Retinopathy Identification Through Deep Learning Methods DOI

Ashok Kumar Kavuru,

Rajesh Kumar Patjoshi

Published: May 2, 2024

Language: Английский

Citations

0
Data-driven 2D-EWT based diabetic retinopathy identification using hybrid neural network DOI
Amit Rawat, Maheshwari Prasad Singh, Rishi Raj Sharma

et al.

Image and Vision Computing, Journal Year: 2024, Volume and Issue: 150, P. 105194 - 105194

Published: July 31, 2024

Language: Английский

Citations

0
Pavement Binder Course Optimization DOI Creative Commons

Abrar Alazawi,

Hassan Musa Al-Mousawi

Al-Nahrain Journal for Engineering Sciences, Journal Year: 2024, Volume and Issue: 27(2), P. 155 - 163

Published: June 20, 2024

In order to avoid losing sense of sight in a large portion the working population, Diabetic Retinopathy (DR) identification during broad examination for diabetes is crucial. To prevent blindness future, early illness detection and measurement disease development are essential. DR diagnosed through medical image analysis. After success Deep Learning (DL) other applications real world, it considered vital tool upcoming health sector applications, providing solutions with accurate results This review provides comprehensive survey state-of-the-art DL models grading using retinal fundus photography. thoroughly examined summarized 81 relevant publications that were published IEEE Xplore, Web Science, PubMed, Scopus between 2018 2023 based on available database binary or multiclass CNN classification as well main preprocessing techniques. According findings this review, transfer learning has proven be an excellent technique addressing problems limited resources data having tens hundreds layers most frequently utilized frameworks classification. The extensively datasets categorization Aptos 2019 EyePACS. Although attained surpassed human-level accuracy, there still more work done real-world clinical procedures.

Language: Английский

Citations

0