Classification of Glaucoma Based on Elephant-Herding Optimization Algorithm and Deep Belief Network

• DOI: 10.3390/electronics11111763
• Publication Date: 2022-06-02T08:46:23Z
• Authors: Mona A. S. Ali; Kishore Balasubra...; Gayathri Devi Kri...; Suresh Muthusamy; Santhiya Pandiyan; Hitesh Panchal; Suman Mann; Kokilavani Thangaraj; Noha E. El-Attar; Laith Abualigah; Diaa Salama Abd E...
• Keywords: Deep belief network; Discriminative model

This study proposes a novel glaucoma identification system from fundus images through the deep belief network (DBN) optimized by elephant-herding optimization (EHO) algorithm. Initially, input image undergoes preprocessing steps of noise removal and enhancement processes, followed optical disc (OD) cup (OC) segmentation extraction structural, intensity, textural features. Most discriminative features are then selected using ReliefF algorithm passed to DBN for classification into glaucomatous or normal. To enhance rate DBN, parameters fine-tuned EHO The model has experimented on public private datasets with 7280 images, which attained maximum 99.4%, 100% specificity, 99.89% sensitivity. 10-fold cross validation reduced misclassification 98.5% accuracy. Investigations proved efficacy proposed method in avoiding bias, dataset variability, reducing false positives compared similar works classification. can be tested diverse datasets, aiding improved diagnosis.