Qiong Song, Xiaofang Huang, Yamin Hu, Kaiqin She
From 254 records retrieved, 7 studies met the inclusion criteria (2023-2026). Applications spanned emergency triage and primary diagnosis, surgical nursing assistance (including one pre-clinical prototype without direct nurse involvement), day ward workflow management, and nursing education. Reported evaluation domains included triage or diagnostic performance, engineering detection performance, usability and acceptability, workflow efficiency, and educational outcomes. AI paradigms included classical machine learning, deep learning (convolutional neural networks), and large language models, deployed via clinical workflow interfaces, augmented reality, and robotic systems, and processing structured and multimodal data (image, text, audio).
INTRODUCTION: Artificial intelligence (AI) is advancing rapidly across clinical medicine, yet its application within ophthalmic nursing, spanning emergency triage, perioperative care, day surgery, and patient education, remains poorly characterized. The growing burden of eye disease and the demand for community-based ophthalmic screening create an urgent context for understanding where AI can augment nursing capacity in this field and what barriers may be limiting evidence generation. This review aimed to systematically map published evidence on AI applications in ophthalmic nursing practice and education, characterize the technologies deployed and outcomes reported, and identify recurring research gaps and barriers through interpretive synthesis of the mapped literature.
METHODS: Four electronic databases (Web of Science Core Collections, PubMed, Scopus, and CINAHL) were searched from inception to June 12, 2026. Peer-reviewed English-language articles simultaneously addressing AI or machine learning, ophthalmology or eye conditions, and nurse engagement were eligible. Records were independently screened by two reviewers and data were extracted using a standardized charting form.
RESULTS: From 254 records retrieved, 7 studies met the inclusion criteria (2023-2026). Applications spanned emergency triage and primary diagnosis, surgical nursing assistance (including one pre-clinical prototype without direct nurse involvement), day ward workflow management, and nursing education. Reported evaluation domains included triage or diagnostic performance, engineering detection performance, usability and acceptability, workflow efficiency, and educational outcomes. AI paradigms included classical machine learning, deep learning (convolutional neural networks), and large language models, deployed via clinical workflow interfaces, augmented reality, and robotic systems, and processing structured and multimodal data (image, text, audio).
DISCUSSION: Despite ophthalmology's prominent role in the early development of clinical AI, dedicated research on AI in ophthalmic nursing remains uncommon in the published literature. To our knowledge, this is the first scoping review to draw on international, English-language sources on this topic. Future research should prioritize nurse-led studies in community and primary care settings, broader modality exploration matched to nursing tasks, rigorous prospective evaluation of nursing-specific outcomes, and, where no current study exists, nurses as active co-designers of AI tools.