RIM-ONE DL: A Unified Retinal Image Database for Assessing Glaucoma Using Deep Learning

Batista, Francisco José Fumero; Díaz-Alemán, Tinguaro; Sigut, J.; Alayón, Silvia; Arnay, Rafael; Ángel-Pereira, Denisse

doi:10.5566/ias.2346

Cited by 64 publications

(32 citation statements)

References 12 publications

(16 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performance of TL-based approaches [32], [34] on the ORIGA dataset is affected by false predictions. Though the approaches' [35]- [37] accuracy is above 93% to 98% on the RIM-ONE dataset, they are suffering from either false positives or negative predictions. On the other side, the PDGC-Net outperforms the currently existing methods.…”

Section: Bulletin Of Electrmentioning

confidence: 95%

Glaucoma classification using a polynomial-driven deep learning approach

Naidana¹,

Barpanda²

2023

Bulletin EEI

View full text Add to dashboard Cite

In this paper, a deep learning-based multi-stage polynomial driven glaucoma classification-net (PDGC-Net) has been proposed for glaucoma identification through retinal images. The proposed approach begins with retinal image pu[1]rification by noise estimation and reduction. Noise has been estimated using a polynomial coefficient-based approach. Images are classified using PDGC-Net, whose polynomial indeterminate representative blocks are designed using new convolutional neural networks (CNN) architectures. The performance of PDGC[1]Net has been observed on the ACRIMA, ORIGA, and retinal image database for optic nerve evaluation (RIM-ONE) datasets. The experimentation is carried out on noisy and denoised images separately, and PDGC-Net has achieved 96% to 98% and 98% to 100% accuracy ranges, respectively. The model’s elasticity is tested with various stages of PDGC-Net. The quantitative PDGC-Net perfor[1]mance analysis is done with state-of-the-art CNN models. The proposed model’s performance has been proven and could be an effective aid to ophthalmologists for glaucoma screening (GS).

show abstract

Section: Bulletin Of Electrmentioning

confidence: 95%

Glaucoma classification using a polynomial-driven deep learning approach

Naidana¹,

Barpanda²

2023

Bulletin EEI

View full text Add to dashboard Cite

show abstract

“…It is optional to train the network with random initial weights, thus reducing the computational cost of training the model. The authors [13], [14], [15], [16], [17] and [18] used pre-trained models to identify glaucoma.…”

Section: Related Workmentioning

confidence: 99%

“…Gómez-Valverde et al [14] explored the application of different CNN architectures to demonstrate the network's performance. Similarly, Batista et al [15], who proposed an improvement in the RIM-ONE dataset, separated the dataset randomly and by the hospital. They performed the extraction and classification with different transfer learning methods in the two proposed separation forms.…”

Section: Related Workmentioning

confidence: 99%

“…Another form of validation was with fifty repetitions in [11], twenty repetitions in [12], and five repetitions in [17]. [15] did not use repetition due to the random and specific separation by the hospital. In [13], repetition was not used.…”

Section: Related Workmentioning

confidence: 99%

“…Therefore, it can return favorable results since identical images could be present in the training and testing data. With dataset improvement in [15], we used the RIM-ONE DL dataset, which contains unduplicated images from the three previous versions of the RIM-ONE dataset.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A Comparative Analysis of Glaucoma Feature Extraction and Classification Techniques in Fundus Images

Assis

Cortez

2023

JCIS

View full text Add to dashboard Cite

Glaucoma is an asymptomatic chronic eye disease that, if not treated in the early stages, can lead to blindness. Therefore, detection in the early stages is essential to preserve the patient’s quality of life. Thus, it is crucial to have a noninvasive method capable of detecting this disease through images in the fundus examination. In the literature, datasets are available with fundus images; however, only a few have glaucoma images and labels. Learning from an imbalanced dataset challenges machine learning, which limits supervised learning algorithms. We compared approaches to extract and classify three public datasets with 2390 images: ACRIMA, REFUGE, and RIM-ONE DL. First, we evaluated extracted features non-structural from HOG, LBP, Zernike, and Gabor filters and features obtained from transfer learning. Then, we classified them with Multilayer Perceptron (MLP), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGB). Finally, each classifier was evaluated individually and in a voting classifier (VOT). We extracted and classified features from transfer learning models in the same process. Also, they were classified using traditional machine learning. Due to class imbalance, we undersampled the majority class normal by applying the following methods: random choice, near miss, and cluster centroid. We also evaluated our model using a cross-dataset approach. Therefore, we efficiently identified glaucoma in different fundus images using network VGG19 and a voting classifier. In addition, balancing classes reduced false negatives and improved model quality. Our approach achieved an average F1-score equals to 94.69%, accuracy rate of 94.77%, precision of 96.10%, recall of 93.45%, and specificity of 96.08%.

show abstract

Deep learning based Glaucoma Network Classification (GNC) using retinal images

Kiyani,

Shehryar,

Khalid

et al. 2023

Int J Imaging Syst Tech

View full text Add to dashboard Cite

The proposed deep learning framework for glaucoma classification addresses critical challenges of limited data and computational costs. Employing data augmentation and normalization techniques, the three‐stage model, utilizing InceptionV3 and ResNet50, achieves high training (99.3% ‐ 99.8%) and testing accuracy (91.6% ‐ 92.12%) on a dataset comprising 16,328 images from fused public datasets. This outperforms existing automated models. The approach leverages transfer learning and convolutional neural networks, showcasing its potential for accurate and timely glaucoma diagnosis. However, ongoing validation on diverse datasets and ethical considerations regarding fairness and transparency in medical applications remain essential. The model's reliability suggests its promising role in aiding early glaucoma detection, potentially averting irreversible vision impairment.

show abstract

RIM-ONE DL: A Unified Retinal Image Database for Assessing Glaucoma Using Deep Learning

Cited by 64 publications

References 12 publications

Glaucoma classification using a polynomial-driven deep learning approach

Glaucoma classification using a polynomial-driven deep learning approach

A Comparative Analysis of Glaucoma Feature Extraction and Classification Techniques in Fundus Images

Deep learning based Glaucoma Network Classification (GNC) using retinal images

Contact Info

Product

Resources

About