Beyond Accuracy: A Controlled Comparative Glaucoma Screening Benchmark Across Deep Learning and Hybrid Models Under Within- and Cross-Dataset Conditions.

Overview

Institutions: Qassim University (Saudi Arabia)

Journal: Journal of clinical medicine, volume 15, issue 16, article 6418

Dates: received 14 July 2026; accepted 16 August 2026; published online 19 August 2026

Type: Research article · Language: English

License: CC BY

Identifiers: DOI 10.3390/jcm15166418 · PMID 42652822 · PMCID PMC13513273 · OpenAlex W7203793722

Open access: gold, a free copy (OpenAlex)

Status: data only

Categories: other (modality), other condition (population), methods / tools (subfield)

Methods: Statistics, Machine learning, Preprocessing, Connectivity, fMRI & imaging

Keywords: glaucoma screening, retinal fundus imaging, deep learning, semi-supervised learning, transfer learning, cross-dataset validation, calibration, benchmark analysis

Topic: Retinal Imaging and Analysis (Radiology, Nuclear Medicine and Imaging, Medicine), according to OpenAlex

Funding: Qassim University (QU-APC-2026)

Citations: not cited yet (Europe PMC); 36 references in the paper

Abstract

Background/Objectives: Artificial intelligence (AI) models for glaucoma screening using colour fundus photography have shown strong internal performance; however, their external validity and calibration reliability remain uncertain. This study developed a controlled four-dataset benchmark to evaluate six glaucoma-screening models across RIM-ONE, DRISHTI-GS, the Hillel Yaffe Glaucoma Dataset, and ORIGA. Methods: The evaluated models included a hybrid deep-handcrafted random forest (RF), transfer-learning and semi-supervised VGG16 models, and a compact convolutional neural network (CNN). Performance was assessed in within-dataset and cross-dataset settings using discrimination metrics, including accuracy, area under the receiver operating characteristic curve (AUC), balanced accuracy, and Matthews correlation coefficient (MCC), as well as calibration metrics, including Brier score and expected calibration error (ECE). Threshold stability, preprocessing ablation, and repeated-seed analyses were also performed. Results: Within-dataset evaluation showed strong discrimination, with the complete hybrid CNN + histogram of oriented gradients (HOG) + local binary patterns (LBP) + minimum redundancy maximum relevance (mRMR) + RF pipeline achieving a mean accuracy of 0.8697 and an AUC of 0.8972. However, cross-dataset performance was poor, with a mean accuracy of 0.5685 and an AUC of 0.5798. The denoising autoencoder-enhanced transfer-learning model showed improved probability calibration in transfer settings. Preprocessing effects were dataset-dependent, with raw images, region of interest (ROI) cropping, and Retinex normalisation producing different external performance patterns. Conclusions: High internal accuracy did not translate into reliable cross-domain generalisation. None of the evaluated models were suitable for zero-shot deployment without site-specific validation or recalibration.

Reproduced under the paper's license (CC BY), from the paper cited above.

Code

The paper links to its data, not to its authors' code: see the Data section.

Tracing map

A tracing map links a paper to the code its authors published: this paper has none, so it has no map.

Data

Datasets cited

doi:10.13026/z0ak-km33, at the source; found in the references

Data Availability Statement

The data presented in this study are available in the article. Further inquiries may be directed to the corresponding author.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 3 authors, 8 keywords, 1 funder, 32 references.

Cite

This paper

Alhasson, H. F., Alharbi, S. S., & Alluwimi, M. S. (2026). Beyond Accuracy: A Controlled Comparative Glaucoma Screening Benchmark Across Deep Learning and Hybrid Models Under Within- and Cross-Dataset Conditions. Journal of clinical medicine, 15(16), 6418. https://doi.org/10.3390/jcm15166418

BibTeX

@article{alhasson2026beyond, author = {Alhasson, Haifa F. and Alharbi, Shuaa S. and Alluwimi, Muhammed S.}, title = {{Beyond Accuracy: A Controlled Comparative Glaucoma Screening Benchmark Across Deep Learning and Hybrid Models Under Within- and Cross-Dataset Conditions}}, journal = {Journal of clinical medicine}, year = {2026}, month = aug, volume = {15}, number = {16}, pages = {6418}, publisher = {Multidisciplinary Digital Publishing Institute (MDPI)}, issn = {2077-0383}, doi = {10.3390/jcm15166418}, url = {https://doi.org/10.3390/jcm15166418}, pmid = {42652822}, pmcid = {PMC13513273} }

RIS

TY - JOUR AU - Alhasson, Haifa F. AU - Alharbi, Shuaa S. AU - Alluwimi, Muhammed S. TI - Beyond Accuracy: A Controlled Comparative Glaucoma Screening Benchmark Across Deep Learning and Hybrid Models Under Within- and Cross-Dataset Conditions T2 - Journal of clinical medicine J2 - J Clin Med PY - 2026 DA - 2026/08/19 VL - 15 IS - 16 SP - 6418 SN - 2077-0383 PB - Multidisciplinary Digital Publishing Institute (MDPI) DO - 10.3390/jcm15166418 UR - https://doi.org/10.3390/jcm15166418 LA - en ER -

CSL-JSON

{ "id": "10.3390/jcm15166418", "type": "article-journal", "title": "Beyond Accuracy: A Controlled Comparative Glaucoma Screening Benchmark Across Deep Learning and Hybrid Models Under Within- and Cross-Dataset Conditions", "container-title": "Journal of clinical medicine", "author": [ { "family": "Alhasson", "given": "Haifa F." }, { "family": "Alharbi", "given": "Shuaa S." }, { "family": "Alluwimi", "given": "Muhammed S." } ], "container-title-short": "J Clin Med", "volume": "15", "issue": "16", "page": "6418", "DOI": "10.3390/jcm15166418", "PMID": "42652822", "PMCID": "PMC13513273", "ISSN": "2077-0383", "publisher": "Multidisciplinary Digital Publishing Institute (MDPI)", "URL": "https://doi.org/10.3390/jcm15166418", "language": "en", "issued": { "date-parts": [ [ 2026, 8, 19 ] ] } }

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.