An Ensemble Framework for Glaucoma Classification Using Fundus Images

Patra, Achirangshu; Nandi, Arijit; Mayaluri, Zefree Lazarus; Lenka, Satyabrata

doi:10.1007/978-981-19-9858-4_49

Cited by 1 publication

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The other five works were using only deep learning models, from which two of the works were combining different types of images as input for the ensemble method [12,14]. We can compare our final results with the most recent work by Patra et al from 2023 [17], because of evaluated results on the same database, REFUGE. Their majority voting ensemble method consisted of individual classifiers with accuracies of 0.9986, 0.9916, and 0.9791 for the Support Vector Machine, Random Forest, and Multilayer Perceptron classifier, respectively.…”

Section: Discussionmentioning

confidence: 98%

Section: Discussionmentioning

confidence: 99%

“…If we compare our work with other authors, who used ensemble methods for glaucoma screening or diagnosis, only one work evaluated their results on the same dataset [17] as we did. Four of the nine works used a combination of deep learning with machine learning classifiers [9][10][11]17], mostly with feature extraction using deep learning models. The other five works were using only deep learning models, from which two of the works were combining different types of images as input for the ensemble method [12,14].…”

Section: Discussionmentioning

confidence: 99%

“…They used the Bagging classifier, AdaBoosting classifier, Random Forest classifier, and Gradient Boosting Classifier, with best accuracy of 90.3% using the Random Forest classifier. The machine learning model "Majority-Voting-Ensemble" consisting of the Support Vector Machine, Random Forest, and Multilayer Perceptron classifiers was proposed by Patra et al [17] with accuracy of 99.86% and 88.48%, on REFUGE and ORIGA datasets, respectively. 13 deep learning models altogether with Support Vector Machine classifier were presented in the ensemble approach in [10].…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Detecting glaucoma from fundus images using ensemble learning

Kurilová,

Rajcsányi,

Rábeková

et al. 2023

Journal of Electrical Engineering

View full text Add to dashboard Cite

Glaucomatous changes of the optic nerve head could be detected from fundus images. Focusing on optic nerve head appearance, and its difference from healthy images, altogether with the availability of plenty of such images in public fundus image databases, these images are ideal sources for artificial intelligence methods applications. In this work, we used ensemble learning methods and compared them with various single CNN models (VGG-16, ResNet-50, and MobileNet). The models were trained on images from REFUGE public dataset. The average voting ensemble method outperformed all mentioned models with 0.98 accuracy. In the AUC metric, the average voting ensemble method outperformed VGG-16 and MobileNet models, which had significantly weaker performance when used alone. The best results were observed using the ResNet-50 model. These results confirmed the significant potential of ensemble learning in enhancing the overall predictive performance in glaucomatous changes detection, but the overall performance could be negatively affected when models with weaker prediction performance are included.

show abstract

Section: Discussionmentioning

confidence: 98%