CNN with Multiple Inputs for Automatic Glaucoma Assessment Using Fundus Images

Elmoufidi, Abdelali; Skouta, Ayoub; Andaloussi, Said Jai; Ouchetto, Ouail

doi:10.1142/s0219467823500122

Cited by 21 publications

(8 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Evaluation on Drishti-GS1 17 revealed that in the case of ResNet50, accuracy remained 0.53 and AUC was 0.75. Rest of the two datasets including RIM-ONE 18,22 sjchoi86-HRF 21 represented accuracy value 0.74 and 0.70, respectively. Shahzad et al fused DenseNet and DarkNet to detect glaucoma at early stages.…”

Section: Comparison Of Deep Learning Models For Glaucoma Classificationmentioning

confidence: 99%

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

Section: Comparison Of Deep Learning Models For Glaucoma Classificationmentioning

confidence: 99%

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

show abstract

“…In most of these previous studies, the CNN models were trained with whole-image CXRs as a single input for the classification [9,11,[13][14][15][16]18,[26][27][28][29][30][31]. Other studies also attempted to develop new CNN models that accept multiple inputs; such multiple-input CNNs (MI-CNN) could effectively improve the classification accuracy and demonstrated better performance than single-input CNNs [32][33][34]. Because an MI-CNN could provide different features, fusing these network features together could improve the accuracy of the entire system [34].…”

Section: Prior Workmentioning

confidence: 99%

Multiple-Inputs Convolutional Neural Network for COVID-19 Classification and Critical Region Screening From Chest X-ray Radiographs: Model Development and Performance Evaluation

Li¹,

Li²,

Yao³

et al. 2022

JMIR Bioinform Biotech

View full text Add to dashboard Cite

Background The COVID-19 pandemic is becoming one of the largest, unprecedented health crises, and chest X-ray radiography (CXR) plays a vital role in diagnosing COVID-19. However, extracting and finding useful image features from CXRs demand a heavy workload for radiologists. Objective The aim of this study was to design a novel multiple-inputs (MI) convolutional neural network (CNN) for the classification of COVID-19 and extraction of critical regions from CXRs. We also investigated the effect of the number of inputs on the performance of our new MI-CNN model. Methods A total of 6205 CXR images (including 3021 COVID-19 CXRs and 3184 normal CXRs) were used to test our MI-CNN models. CXRs could be evenly segmented into different numbers (2, 4, and 16) of individual regions. Each region could individually serve as one of the MI-CNN inputs. The CNN features of these MI-CNN inputs would then be fused for COVID-19 classification. More importantly, the contributions of each CXR region could be evaluated through assessing the number of images that were accurately classified by their corresponding regions in the testing data sets. Results In both the whole-image and left- and right-lung region of interest (LR-ROI) data sets, MI-CNNs demonstrated good efficiency for COVID-19 classification. In particular, MI-CNNs with more inputs (2-, 4-, and 16-input MI-CNNs) had better efficiency in recognizing COVID-19 CXRs than the 1-input CNN. Compared to the whole-image data sets, the efficiency of LR-ROI data sets showed approximately 4% lower accuracy, sensitivity, specificity, and precision (over 91%). In considering the contributions of each region, one of the possible reasons for this reduced performance was that nonlung regions (eg, region 16) provided false-positive contributions to COVID-19 classification. The MI-CNN with the LR-ROI data set could provide a more accurate evaluation of the contribution of each region and COVID-19 classification. Additionally, the right-lung regions had higher contributions to the classification of COVID-19 CXRs, whereas the left-lung regions had higher contributions to identifying normal CXRs. Conclusions Overall, MI-CNNs could achieve higher accuracy with an increasing number of inputs (eg, 16-input MI-CNN). This approach could assist radiologists in identifying COVID-19 CXRs and in screening the critical regions related to COVID-19 classifications.

show abstract

“…The condition is regarded as severe since the damage it causes is irredeemable, leading to perpetual vision loss if not promptly treated. While there is no prophylactic treatment for glaucoma, it is possible to avoid blindness by detecting and curing the condition at an initial stage [9][10][11][12][13][14][15]. Glaucoma is normally diagnosed by measures such as: obtaining a medical history of the patient; measuring intraocular pressure; performing a visual field loss test; and conducting a manual assessment of the optic disc using ophthalmoscopy to examine the shape and color of the optic nerve.…”

Section: Introductionmentioning

confidence: 99%

Creating an Early Diagnostic Method for Glaucoma Using Convolutional Neural Networks

Alqarni,

Al Harbi,

Subhan

2024

Preprint

View full text Add to dashboard Cite

According to the World Health Organization, glaucoma is a leading cause of blindness, accounting for over 12% of global blindness as it affects one in every 100 people. In fact, 79.6 million people worldwide live with blindness caused by glaucoma. This is because the current method for diagnosing glaucoma is by examining retinal fundus images. However, it is considerably difficult to distinguish the lesions' features solely through manual observations by ophthalmologists, especially in the early phases. This study proposes a new diagnosis method using convolutional neural networks. The attention mechanism is utilized to learn pixel-wise features for accurate prediction. Several attention strategies have been developed to guide the networks in learning the important features and factors that affect localization accuracy. The algorithms were trained for glaucoma detection using Python 2.7, TensorFlow, Py Torch, and Keras. The methods were evaluated on Drishti-GS and RIM-ONE datasets with 361 training and 225 test sets, consisting of 344 healthy and 242 glaucomatous images. The proposed algorithms can achieve impressive results that show an increase in overall diagnostic efficiency, as the algorithm displays a 30-second detection time with 98.9% accuracy compared to the 72.3% accuracy of traditional testing methods. Finally, this algorithm has been implemented as a webpage, allowing patients to test for glaucoma. This webpage offers various services such as: connecting the patient to the nearest care setup; offering scientific articles regarding glaucoma; and a video game that supports eye-treatment yogic exercises to strengthen vision and focus. This early diagnostic method has the near future potential to decrease the percentage of irreversible vision loss due to glaucoma by 42.79% (the percentage was calculated using the mean absolute error function), which could prevent glaucoma from remaining the leading cause of blindness worldwide.

show abstract

CNN with Multiple Inputs for Automatic Glaucoma Assessment Using Fundus Images

Cited by 21 publications

References 47 publications

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

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

Multiple-Inputs Convolutional Neural Network for COVID-19 Classification and Critical Region Screening From Chest X-ray Radiographs: Model Development and Performance Evaluation

Creating an Early Diagnostic Method for Glaucoma Using Convolutional Neural Networks

Contact Info

Product

Resources

About