A deep learning based model for diabetic retinopathy grading

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

Published: Jan. 30, 2025

Diabetic retinopathy stands as a leading cause of blindness among people. Manual examination DR images is labor-intensive and prone to error. Existing methods detect this disease often rely on handcrafted features which limit the adaptability classification accuracy. Thus, aim research develop an automated efficient system for early detection accurate grading diabetic severity with less time consumption. In our research, we have developed deep neural network named RSG-Net (Retinopathy Severity Grading) classify into 4 stages (multi-class classification) 2 (binary classification). The dataset utilized in study Messidor-1. preprocessing, used Histogram Equalization improve image contrast denoising techniques remove noise artifacts enhanced clarity fundus images. We applied data augmentation preprocessed order tackle class imbalance issues. Augmentation involve flipping, rotation, zooming adjustment color, brightness. proposed model contains convolutional layers perform automatic feature extraction from input batch normalization training speed performance. also max pooling, drop out fully connected layers. Our achieved testing accuracy 99.36%, specificity 99.79% sensitivity 99.41% classifying grades it 99.37% accuracy, 100% 98.62% grades. performance compared other state-of-the-art methodologies where outperformed these methods.

Language: Английский

---

ResViT FusionNet Model: An explainable AI-driven approach for automated grading of diabetic retinopathy in retinal images

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 186, P. 109656 - 109656

Published: Jan. 16, 2025

Language: Английский

---

Enhanced multi-grade diabetic retinopathy detection and classification via ensembled deep learning model from retinal fundus images

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 128116 - 128116

Published: May 1, 2025

Language: Английский

---

A Customized CNN Architecture with CLAHE for Multi-Stage Diabetic Retinopathy Classification

Engineering Technology & Applied Science Research, Journal Year: 2024, Volume and Issue: 14(6), P. 18258 - 18263

Published: Dec. 2, 2024

This paper presents a customized Convolutional Neural Network (CNN) architecture for multi-stage detection of Diabetic Retinopathy (DR), leading cause vision impairment and blindness. The proposed model incorporates advanced image enhancement techniques, particularly Contrast Limited Adaptive Histogram Equalization (CLAHE), to improve the visibility critical retinal features associated with DR. By integrating CLAHE finely tuned CNN, approach significantly enhances accuracy robustness, allowing more precise across various stages was evaluated against several state-of-the-art CNN alone achieving an overall 97.69%. addition further boosts performance, 99.69%, underscoring effectiveness combining automated DR detection. provides efficient, scalable, highly accurate solution early multistage detection, which is crucial timely intervention prevention loss.

Language: Английский

---

Diabetic Retinopathy Prediction Based on a Hybrid Deep Learning Approach

Published: May 8, 2024

Language: Английский

---

Deep Learning for the Detection and Classification of Diabetic Retinopathy Stages

Microsystems Electronics and Acoustics, Journal Year: 2024, Volume and Issue: 29(2)

Published: Aug. 4, 2024

The incidence of diabetic retinopathy (DR), a complication diabetes leading to severe vision impairment and potential blindness, has surged worldwide in recent years. This condition is considered one the causes loss. To improve diagnostic accuracy for DR reduce burden on healthcare professionals, artificial intelligence (AI) methods are increasingly implemented medical institutions. AI-based models, particular, integrating more algorithms enhance performance existing neural network architectures that commercially used detection. However, these models still exhibit limitations, such as need high computational power lower detecting early stages. overcome challenges, developing advanced machine learning precise detection classification stages essential, it would aid ophthalmologists making accurate diagnoses. article reviews current research use deep diagnosing classifying related diseases, well challenges face this solutions early-stage review provides information modern approaches using applications discusses issues limitations area.

Language: Английский

---