Artificial intelligence in oculoplastic surgery: A Preferred Reporting Items for Systematic Reviews and Meta-Analyses-based systematic review and meta-analysis of diagnostic accuracy, surgical planning, and clinical integration
Journal of Clinical Ophthalmology and Research · 2026-10 · 14 (4) · 453-465
DOI: 10.4103/jcor.jcor_101_26 ↗Kurz gefasst
Systematische Übersicht in der Zeitschrift Journal of Clinical Ophthalmology and Research, online erschienen am 18.11.2026. Verfasst von Niraj Kumar Yadav und 4 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Background
Artificial intelligence (AI) is revolutionizing image-driven medical specialties. Oculoplastic and orbital surgery depends on clinical assessment as well as surgical planning that are based on visual interpretation, anatomical precision, and millimeter-level measurements and thus may benefit greatly from this.
Objective
The objective of this study was to review the clinical applications, diagnostic performance, surgical utility, methodological quality, and translational readiness of standard AI systems in oculoplastic and orbital surgery.
Methods
This systematic review and meta-analysis has been reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines and was prospectively registered in PROSPERO (CRD420261433664). PubMed/MEDLINE, Scopus, Web of Science, and Embase were searched for studies published between January 2000 and December 2025. Eligible studies assessed AI systems for periocular, eyelid, lacrimal, or orbital diseases. Evidence was synthesized qualitatively, or, where appropriate, quantitatively using random effects models. Diagnostic subgroup analyses included classical machine learning, convolutional neural network (CNN)-based deep learning, and transformer or hybrid architectures.
Results
Of the 166 included studies, 92 were available for meta-analysis. Diagnostic pooled area under the receiver-operating characteristic curve (AUC) was 0.93 (95% confidence interval [CI], 0.91–0.95), sensitivity was 0.90 (95% CI, 0.87–0.93), and specificity was 0.88 (95% CI, 0.85–0.91). CNN-based models outperformed classical machine learning algorithms (summary AUC, 0.94 vs. 0.88). The Dice coefficient for orbital segmentation was 0.89 pooled. Predictive models could predict postoperative eyelid height and proptosis to within submillimeter accuracy. Tremor reduction and real-time tracking were provided by intraoperative systems, and postoperative platforms demonstrated high agreement with surgeon assessment and complication identification.
Conclusion
AI shows good diagnostic accuracy and potential perioperative use in oculoplastic surgery. The emerging applications are three-dimensional reconstruction, virtual surgical planning and smartphone-based monitoring. However, heterogeneity, limited external validation and potential dataset bias warrant prospective multicenter validation, standardized reporting, secure governance and clinician oversight before widespread implementation.
Zitieren
APA
Yadav, N. K., Singhal, S., Srivastava, S., Wadhwani, M., & Hada, M. (2026). Artificial intelligence in oculoplastic surgery: A Preferred Reporting Items for Systematic Reviews and Meta-Analyses-based systematic review and meta-analysis of diagnostic accuracy, surgical planning, and clinical integration. Journal of Clinical Ophthalmology and Research, 14(4), 453-465. https://doi.org/10.4103/jcor.jcor_101_26
Vancouver
Yadav NK, Singhal S, Srivastava S, Wadhwani M, Hada M. Artificial intelligence in oculoplastic surgery: A Preferred Reporting Items for Systematic Reviews and Meta-Analyses-based systematic review and meta-analysis of diagnostic accuracy, surgical planning, and clinical integration. Journal of Clinical Ophthalmology and Research. 2026;14(4):453-465. doi:10.4103/jcor.jcor_101_26
BibTeX
@article{Yadav2026,
author = {Yadav, Niraj Kumar and Singhal, Semone and Srivastava, Shachi and Wadhwani, Meenakshi and Hada, Maya},
title = {Artificial intelligence in oculoplastic surgery: A Preferred Reporting Items for Systematic Reviews and Meta-Analyses-based systematic review and meta-analysis of diagnostic accuracy, surgical planning, and clinical integration},
journal = {Journal of Clinical Ophthalmology and Research},
year = {2026},
volume = {14},
number = {4},
pages = {453--465},
doi = {10.4103/jcor.jcor_101_26}
}
EUVIMED