STAGE challenge: Structural-Functional Transition in Glaucoma Assessment

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not reported in abstract; specific performance metrics (e.g., mean absolute error, correlation coefficients, Dice scores) not provided

• Primary Outcome: Prediction of functional visual field indicators (mean deviation, sensitivity maps, and pattern deviation probability maps) from structural OCT volumes using deep learning; all seven finalist teams employed dual-branch architectures integrating OCT with tabular demographic data

• Nnt Or Sensitivity: Not applicable in standard form; diagnostic/predictive accuracy metrics not reported in abstract. Teams using task-specific OCT models qualitatively reported to achieve highest performance, but no sensitivity, specificity, or AUC values are provided

• Confidence Interval: Not reported

Clinical Application

Currently a research-stage technology. Clinical deployment would require: (1) external validation across diverse populations and OCT platforms; (2) regulatory approval; (3) integration into existing ophthalmic imaging workflows; (4) clinician training and acceptance; (5) demonstration of non-inferiority to standard VF testing for clinical decision-making. The 401-case dataset, while useful for benchmarking, is insufficient for regulatory-grade validation. In Australia, glaucoma affects approximately 300,000 people, with significant under-diagnosis in the community. Visual field testing is a Medicare Benefits Schedule (MBS) item (items 10940–10942) and is central to glaucoma monitoring under RANZCO and Optometry Australia guidelines. The TGA would require prospective clinical validation before any such predictive tool could be marketed as a medical device (Software as a Medical Device, SaMD, under TGA's Digital Health framework). The RACGP's Red Book does not currently recommend OCT-based VF prediction tools in primary care. This technology could have particular relevance for rural and remote Australian communities where access to VF testing is limited, but robust local validation data are absent. PBS implications are not directly relevant at this stage. Adults with established or suspected glaucoma undergoing routine OCT imaging in ophthalmology or optometry settings, particularly where access to reliable VF testing is limited by patient cooperation, time constraints, or resource availability

Abstract

Glaucoma is a leading cause of irreversible yet preventable blindness in working-age populations. Clinical diagnos is currently relies on functional visual field (VF) examinations to evaluate visual function and monitor disease progression, but these tests are time-consuming and require close cooperation with ophthalmologists, limiting their practicality. Although deep learning has shown promise for glaucoma diagnosis, most models have focused on structural changes in fundus and OCT images without linking them to functional VF outcomes. To bridge this gap, the HDMI Laboratory, in collaboration with the Zhongshan Ophthalmic Center of Sun Yat-sen University, organized the STAGE Challenge: Structural-Functional Transition in Glaucoma Assessment. The challenge explores predicting functional VF indicators-mean deviation, sensitivity maps, and pattern deviation probability maps-directly from structural OCT images. A dataset of 401 OCT volumes (each with 256 cross-sectional images) is released with corresponding VF labels and demographic data, along with a standardized evaluation framework to ensure fair comparison. This paper summarizes the methods of the seven finalist teams and analyzes their results. All teams employed dual-branch architectures integrating OCT with tabular data, and those using task-specific OCT models achieved the highest performance. These findings highlight the importance of tailored deep learning strategies for linking structural imaging to functional outcomes. The STAGE Challenge thus establishes the first standardized benchmark with a large, curated dataset for this task, enabling systematic evaluation of algorithms for structure-function analysis in glaucoma, and providing a foundation for future research. Details of the competition are available at http://hdmilab.cn/competition/stage.

References

1. 1.Zhou, S., Liang, Y., Fang, H., Liu, D., Wei, L., Chen, Z., Xia, Y., Ou, C., Li, Y., Tan, Y., Yang, B., Yin, H., Feng, C., Zhou, H., Bogunović, H., Fu, H., Li, F., Zhang, X., & Xu, Y. (2026). STAGE challenge: Structural-Functional Transition in Glaucoma Assessment. Medical Image Analysis. https://doi.org/10.1016/j.media.2026.104128