Explainable Hybrid CNN-SVM Framework with SHAP-based Recursive Feature Elimination for Glaucoma Classification

Read full text Summarise with VIRAT
Authors: Fingki Sri Marsanti, Agus Harjoko
Cited by: 0
References: 28

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Glaucoma represent a primary contributor to irreversible blindness worldwide, with approximately 3.61 million cases documented in 2020.Its elevated incidence stems from the absence of early symptoms, resulting in deferred diagnosis and intervention.Traditional diagnostic approaches reliant on manual fundus image assessment are inherently subjective and contingent upon clinical proficiency.This research proposed an automated glaucoma classification framework integrating Convolutional Neural Network (CNN) and Support Vector Machine (SVM), enhanced through SHAP-based Recursive Feature Elimination (RFE-SHAP) for feature selection.Evaluated on the ACRIMA and ORIGA datasets, the framework prioritizes both performance and interpretability.Optimal outcomes were observed on ACRIMA, where SVM employing selected DenseNet-201 features attained an accuracy 97.87%, precision 96.34%, and recall 100%.Conversely, diminished performance on ORIGA underscores dataset heterogeneity impacting model generalization.Furthermore, SHAP-based interpretability analysis successfully explains the contribution of each feature, supporting the development of transparent and explainable AI-driven diagnostic systems.

Fields of Study & Subjects

Retinal Imaging and Analysis Glaucoma and retinal disorders Ophthalmology and Visual Health Research

Publication details

Journal / source
International journal of intelligent engineering and systems
Volume
19
Issue
4
Pages
737–753
Publication date
23 March 2026
Language
EN