ResDenseNet:Hybrid Convolutional Neural Network Model for Advanced Classification of Diabetic Retinopathy(DR) in Retinal Image Analysis

International Journal of Computational and Experimental Science and Engineering, Journal Year: 2024, Volume and Issue: 10(4)

Published: Dec. 22, 2024

Preventing vision loss in diabetic retinopathy (DR) requires early and precise detection. Although strong feature extraction is required there class imbalance the current methods, deep learning (DL) techniques have showed promise DR classification. With components from both ResNeXt DenseNet designs, a unique DL architecture for classification proposed this work. A that integrates work.To address issues classification, method channel-wise masking with an attention mechanism. The network able to learn less frequent stages because reduces influence of majority concentrates on important features. To improve interpretability confidence model's predictions, incorporation Explainable AI (XAI) approaches also covered.Our findings show suggested approach outperforms architectures, achieving better sensitivity differentiating phases at 0.82 accuracy 0.87. This shows new has improving categorization, which could result earlier diagnoses patient outcomes.

Language: Английский

---

Identification of diabetic retinopathy lesions in fundus images by integrating CNN and vision mamba models

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(1), P. e0318264 - e0318264

Published: Jan. 28, 2025

Diabetic retinopathy, a retinal disorder resulting from diabetes mellitus, is prominent cause of visual degradation and loss among the global population. Therefore, identification classification diabetic retinopathy are utmost importance in clinical diagnosis therapy. Currently, these duties extensively carried out by manual examination utilizing human system. Nevertheless, sometimes arduous, time-consuming, prone to errors. Deep learning-based methods have recently demonstrated encouraging results several areas, such as image categorization natural language mining. The majority deep learning techniques developed for medical analysis rely on convolutional modules extract inherent structure images within certain local receptive field. Furthermore, transformer-based models been utilized tackle processing problems capitalizing connections distant pixels images. Considering analyses, this work presents comprehensive model that combines neural network vision mamba models. This designed accurately identify classify lesions displayed fundus component incorporates bidirectional state space method positional embedding enable location sensitivity data samples meet conditions relationship context. An evaluation suggested was comparison experiments between state-of-the-art algorithms proposed methodology. Empirical findings demonstrate methodology surpasses most advanced datasets accessible openly. Hence, approach may be regarded helpful tool therapeutic processes.

Language: Английский

---

PHOTODIAGNOSIS WITH DEEP LEARNING: A GAN AND AUTOENCODER-BASED APPROACH FOR DIABETIC RETINOPATHY DETECTION

Photodiagnosis and Photodynamic Therapy, Journal Year: 2025, Volume and Issue: unknown, P. 104552 - 104552

Published: March 1, 2025

Diabetic retinopathy (DR) is a leading cause of visual impairment and blindness worldwide, necessitating early detection accurate diagnosis. This study proposes novel framework integrating Generative Adversarial Networks (GANs) for data augmentation, denoising autoencoders noise reduction, transfer learning with EfficientNetB0 to enhance the performance DR classification models. GANs were employed generate high-quality synthetic retinal images, effectively addressing class imbalance enriching training dataset. Denoising further improved image quality by reducing eliminating common artifacts such as speckle noise, motion blur, illumination inconsistencies, providing clean consistent inputs model. was fine-tuned on augmented denoised The achieved exceptional metrics, including 99.00% accuracy, recall, specificity, surpassing state-of-the-art methods. custom-curated OCT dataset featuring high-resolution clinically relevant challenges limited annotated noisy inputs. Unlike existing studies, our work uniquely integrates GANs, autoencoders, EfficientNetB0, demonstrating robustness, scalability, clinical potential proposed framework. Future directions include interpretability tools adoption exploring additional imaging modalities improve generalizability. highlights transformative deep in critical diabetic

Language: Английский

---

Multi-model approach for precise lesion localization and severity grading for diabetic retinopathy and age-related macular degeneration

Frontiers in Computer Science, Journal Year: 2025, Volume and Issue: 7

Published: April 15, 2025

Introduction Accurate and efficient automated diagnosis of Diabetic Retinopathy (DR) Age-related Macular Degeneration (AMD) is crucial for addressing these leading causes vision loss worldwide. Driven by the potential to improve early detection patient outcomes, this study proposes a comprehensive system diagnosing grading conditions. Methods Our approach combines image enhancement techniques, lesion localization, disease severity classification. The utilizes both established benchmark datasets four newly proposed ensure robust evaluation. Results localization model achieved exceptional performance with mAP scores up 98.71% AMD on Shiromoni_AMD dataset 97.21% DR KLC_DR dataset. Similarly, classification demonstrated high accuracy, reaching 99.42% Stare 98.81% Comparative analysis shows that our methods often surpass existing state-of-the-art approaches, demonstrating more consistent across diverse eye Discussion This research represents significant advancement in ophthalmic diagnosis, potentially enhancing clinical practice improving accessibility care findings pave way accurate, efficient, widely applicable screening tools retinal diseases.

Language: Английский

---

Automated diabetic retinopathy detection using a multi-step framework with stacked ensemble-based classification model

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127709 - 127709

Published: April 1, 2025

Language: Английский

---

EfficientNetB0-Based Automated Diabetic Retinopathy Classification in Fundus Images

IGI Global eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 311 - 338

Published: May 2, 2025

The most common cause of blindness among diabetics is diabetic retinopathy. This disease must be detected in its early stages as delaying treatment can result permanent blindness. In today's modern world with the development technologies, automated techniques have improved accuracy and efficiency detecting classifying this disease. study suggests a deep learning model utilizing EfficientNetB0 that focuses on designing neural network structure requires fewer parameters computations compared to traditional architectures. Further help increasing accuracy, image being processed methods like CLAHE which an processing algorithm while keeping noise images low possible. Also, augmentation done SMOTE applied handle class imbalance. We collected APTOS dataset consisting retinal fundus images. obtained 98.78% indicating model's effectiveness between various

Language: Английский

---

Real-Time Diabetic Retinopathy Detection Using YOLO-v10 with Nature-Inspired Optimization

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: May 7, 2025

Language: Английский

---

Enhancing Diabetic Retinopathy Classification Using Feature Extraction Algorithm

Published: Jan. 1, 2025

Language: Английский

---

Creating an Automatic Classifier for Detecting Diabetic Retinopathy Based on the Combining a New 7-Class Dataset Fundus Images and a Collection of Images from Another Source

Lecture notes in networks and systems, Journal Year: 2025, Volume and Issue: unknown, P. 246 - 255

Published: Jan. 1, 2025

Language: Английский

---

Detection of Diabetic Retinopathy Using Discrete Wavelet-Based Center-Symmetric Local Binary Pattern and Statistical Features

Deleted Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 5, 2024

Computer-aided diagnosis (CAD) system assists ophthalmologists in early diabetic retinopathy (DR) detection by automating the analysis of retinal images, enabling timely intervention and treatment. This paper introduces a novel CAD based on global multi-resolution images. As first step, we enhance quality images applying sequence preprocessing techniques, which include median filter, contrast limited adaptive histogram equalization (CLAHE), unsharp filter. These steps effectively eliminate noise Further, these are represented at multi-scales using discrete wavelet transform (DWT), center symmetric local binary pattern (CSLBP) features extracted from each scale. The CSLBP decomposed capture fine coarse details fundus Also, statistical to characteristics provide comprehensive representation performances evaluated benchmark dataset two machine learning models, i.e., SVM k-NN, found that performance proposed work is considerably more encouraging than other existing methods. Furthermore, results demonstrate when wavelet-based combined with features, they yield notably improved compared individually.

Language: Английский

---