A Generative AI–Enhanced Multimodal Deep Learning Framework for Early Glaucoma Detection using Retinal Imaging and Clinical Features
Year: 2026
Type: conference-paper
Abstract: Recent ophthalmic research and studies concludes the diagnosis of glaucoma in early stages is more difficult and hard. In same time, if the diagnosis is not done and failed to predict the glaucoma may leads to unnoticed structural changes in the retina may cause the vision disability. One of the research confirms that machines can assist the predicting process in glaucoma in starting stages. The Learning process of Eye scans uses advanced pattern recognition methods and algorithms. Most of the existing systems rely on single type scanning scheme. Many diagnostic models are using the limited labelled data’s for further processing. This kind of dependency may reduce the performance over different group of patients . In the procedure of reliable diagnosis, identification of optic cup regions and disc plays main portion. The consistency of diagnostic outcomes rely on enhanced segmentation process. Moreover the integration of number of scan types provide better complete diagnostic view. The clinical information’s and layered OCT data are combined with raised resolution fundus images gives better level in diagnostics. To improve data quality and segmentation clarity, Gen-AI methods may helpful. Another research concludes that the synthetic data improvement methods will avoid data scarcity. Also another finding shows much advantages on combining clinical features and imaging [. During the diagnostic process, if the data sets are varied, leads to better accuracy. Here most of the research and findings shows the major requirement of glaucoma screening process need to be improved and must be reliable in real clinical environments.
Authors: Mary Anitha T, Ramalakshmi K.
Institution: Alliance University
Language: English
Topic: Retinal Imaging and Analysis
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7204140384