Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fair-Eye Net: A Fair, Trustworthy, Multimodal Integrated Glaucoma Full Chain AI System

Wenbin Wei, Suyuan Yao, Cheng Huang, Xiangyu Gao

TL;DR

Fair-Eye Net addresses the fragmented and biased nature of glaucoma care by proposing a multimodal, uncertainty-aware AI system that jointly performs screening, prognosis, and dynamic risk warning. It introduces a dual-stream architecture with uncertainty-based gating and multi-task optimization to deliver accurate results while calibrating for fairness across diverse populations. Experimental results show strong discriminative performance (AUC ≈ 0.912) with high specificity, substantially reduced racial FNR gaps, and lead times for risk alerts, supporting safe clinical deployment. The framework emphasizes robustness to cross-device/domain shifts and treats fairness as a primary design objective, offering a scalable path toward equitable, longitudinal glaucoma management.

Abstract

Glaucoma is a top cause of irreversible blindness globally, making early detection and longitudinal follow-up pivotal to preventing permanent vision loss. Current screening and progression assessment, however, rely on single tests or loosely linked examinations, introducing subjectivity and fragmented care. Limited access to high-quality imaging tools and specialist expertise further compromises consistency and equity in real-world use. To address these gaps, we developed Fair-Eye Net, a fair, reliable multimodal AI system closing the clinical loop from glaucoma screening to follow-up and risk alerting. It integrates fundus photos, OCT structural metrics, VF functional indices, and demographic factors via a dual-stream heterogeneous fusion architecture, with an uncertainty-aware hierarchical gating strategy for selective prediction and safe referral. A fairness constraint reduces missed diagnoses in disadvantaged subgroups. Experimental results show it achieved an AUC of 0.912 (96.7% specificity), cut racial false-negativity disparity by 73.4% (12.31% to 3.28%), maintained stable cross-domain performance, and enabled 3-12 months of early risk alerts (92% sensitivity, 88% specificity). Unlike post hoc fairness adjustments, Fair-Eye Net optimizes fairness as a primary goal with clinical reliability via multitask learning, offering a reproducible path for clinical translation and large-scale deployment to advance global eye health equity.

Fair-Eye Net: A Fair, Trustworthy, Multimodal Integrated Glaucoma Full Chain AI System

TL;DR

Fair-Eye Net addresses the fragmented and biased nature of glaucoma care by proposing a multimodal, uncertainty-aware AI system that jointly performs screening, prognosis, and dynamic risk warning. It introduces a dual-stream architecture with uncertainty-based gating and multi-task optimization to deliver accurate results while calibrating for fairness across diverse populations. Experimental results show strong discriminative performance (AUC ≈ 0.912) with high specificity, substantially reduced racial FNR gaps, and lead times for risk alerts, supporting safe clinical deployment. The framework emphasizes robustness to cross-device/domain shifts and treats fairness as a primary design objective, offering a scalable path toward equitable, longitudinal glaucoma management.

Abstract

Glaucoma is a top cause of irreversible blindness globally, making early detection and longitudinal follow-up pivotal to preventing permanent vision loss. Current screening and progression assessment, however, rely on single tests or loosely linked examinations, introducing subjectivity and fragmented care. Limited access to high-quality imaging tools and specialist expertise further compromises consistency and equity in real-world use. To address these gaps, we developed Fair-Eye Net, a fair, reliable multimodal AI system closing the clinical loop from glaucoma screening to follow-up and risk alerting. It integrates fundus photos, OCT structural metrics, VF functional indices, and demographic factors via a dual-stream heterogeneous fusion architecture, with an uncertainty-aware hierarchical gating strategy for selective prediction and safe referral. A fairness constraint reduces missed diagnoses in disadvantaged subgroups. Experimental results show it achieved an AUC of 0.912 (96.7% specificity), cut racial false-negativity disparity by 73.4% (12.31% to 3.28%), maintained stable cross-domain performance, and enabled 3-12 months of early risk alerts (92% sensitivity, 88% specificity). Unlike post hoc fairness adjustments, Fair-Eye Net optimizes fairness as a primary goal with clinical reliability via multitask learning, offering a reproducible path for clinical translation and large-scale deployment to advance global eye health equity.
Paper Structure (22 sections, 5 equations, 6 figures, 3 tables)

This paper contains 22 sections, 5 equations, 6 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. The Fair-eye Framework
  4. Dual-Stream Heterogeneous Architecture
  5. Uncertainty-Aware Hierarchical Gating
  6. Multi-task Synergistic Optimization
  7. Experimental Design
  8. Datasets: A Hierarchical Validation Ecosystem
  9. Data Preprocessing Pipeline
  10. Implementation Details
  11. Experimental Results and Discussion
  12. Overview of Results
  13. Core Performance Analysis
  14. Integrated Outputs: Interpretability and Severity Quantification
  15. Screening Performance: Trade-offs and Advantages vs. SOTA/Baselines
  16. ...and 7 more sections

Figures (6)

  • Figure 1: The Proposed Fair-Eye Framework.
  • Figure 2: Glaucoma severity analysis.
  • Figure 3: Comparison of Fair-Eye Net with SOTA methods and baselines.
  • Figure 4: Overall results of screening, risk prediction, and dynamic warning in Fair-Eye Net.
  • Figure 5: Coverage-accuracy curve.
  • ...and 1 more figures