Saved in:
Bibliographic Details
Main Authors: Kar, Susmita, Akib, A S M Ahsanul Sarkar, Hasib, Abdul, Yaser, Samin, Azim, Anas Bin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.08240
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915725903396864
author Kar, Susmita
Akib, A S M Ahsanul Sarkar
Hasib, Abdul
Yaser, Samin
Azim, Anas Bin
author_facet Kar, Susmita
Akib, A S M Ahsanul Sarkar
Hasib, Abdul
Yaser, Samin
Azim, Anas Bin
contents Diabetic retinopathy (DR), affecting millions globally with projections indicating a significant rise, poses a severe blindness risk and strains healthcare systems. Diagnostic complexity arises from visual symptom overlap with conditions like age-related macular degeneration and hypertensive retinopathy, exacerbated by high misdiagnosis rates in underserved regions. This study introduces TIMM-ProRS, a novel deep learning framework integrating Vision Transformer (ViT), Convolutional Neural Network (CNN), and Graph Neural Network (GNN) with multi-modal fusion. TIMM-ProRS uniquely leverages both retinal images and temporal biomarkers (HbA1c, retinal thickness) to capture multi-modal and temporal dynamics. Evaluated comprehensively across diverse datasets including APTOS 2019 (trained), Messidor-2, RFMiD, EyePACS, and Messidor-1 (validated), the model achieves 97.8\% accuracy and an F1-score of 0.96, demonstrating state-of-the-art performance and outperforming existing methods like RSG-Net and DeepDR. This approach enables early, precise, and interpretable diagnosis, supporting scalable telemedical management and enhancing global eye health sustainability.
format Preprint
id arxiv_https___arxiv_org_abs_2601_08240
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Temporal-Enhanced Interpretable Multi-Modal Prognosis and Risk Stratification Framework for Diabetic Retinopathy (TIMM-ProRS)
Kar, Susmita
Akib, A S M Ahsanul Sarkar
Hasib, Abdul
Yaser, Samin
Azim, Anas Bin
Image and Video Processing
Computer Vision and Pattern Recognition
Diabetic retinopathy (DR), affecting millions globally with projections indicating a significant rise, poses a severe blindness risk and strains healthcare systems. Diagnostic complexity arises from visual symptom overlap with conditions like age-related macular degeneration and hypertensive retinopathy, exacerbated by high misdiagnosis rates in underserved regions. This study introduces TIMM-ProRS, a novel deep learning framework integrating Vision Transformer (ViT), Convolutional Neural Network (CNN), and Graph Neural Network (GNN) with multi-modal fusion. TIMM-ProRS uniquely leverages both retinal images and temporal biomarkers (HbA1c, retinal thickness) to capture multi-modal and temporal dynamics. Evaluated comprehensively across diverse datasets including APTOS 2019 (trained), Messidor-2, RFMiD, EyePACS, and Messidor-1 (validated), the model achieves 97.8\% accuracy and an F1-score of 0.96, demonstrating state-of-the-art performance and outperforming existing methods like RSG-Net and DeepDR. This approach enables early, precise, and interpretable diagnosis, supporting scalable telemedical management and enhancing global eye health sustainability.
title Temporal-Enhanced Interpretable Multi-Modal Prognosis and Risk Stratification Framework for Diabetic Retinopathy (TIMM-ProRS)
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.08240