Applications of Artificial Intelligence in Cataract Surgery: A Review.

Autor:	Ahuja AS; Department of Ophthalmology, Casey Eye Institute, Oregon Health and Science University, Portland, OR, USA., Paredes Iii AA; Charles E. Schmidt College of Medicine, Florida Atlantic University, Boca Raton, FL, USA., Eisel MLS; College of Medicine, Florida State University, Tallahassee, FL, USA., Kodwani S; Windsor University School of Medicine, Cayon, St. Kitts, KN., Wagner IV; Department of Ophthalmology, Mayo Clinic Florida, Jacksonville, FL, USA., Miller DD; Department of Ophthalmology, Mayo Clinic Florida, Jacksonville, FL, USA., Dorairaj S; Department of Ophthalmology, Mayo Clinic Florida, Jacksonville, FL, USA.
Jazyk:	angličtina
Zdroj:	Clinical ophthalmology (Auckland, N.Z.) [Clin Ophthalmol] 2024 Oct 17; Vol. 18, pp. 2969-2975. Date of Electronic Publication: 2024 Oct 17 (Print Publication: 2024).
DOI:	10.2147/OPTH.S489054
Abstrakt:	Cataract surgery is one of the most performed procedures worldwide, and cataracts are rising in prevalence in our aging population. With the increasing utilization of artificial intelligence (AI) in the medical field, we aimed to understand the extent of present AI applications in ophthalmic microsurgery, specifically cataract surgery. We conducted a literature search on PubMed and Google Scholar using keywords related to the application of AI in cataract surgery and included relevant articles published since 2010 in our review. The literature search yielded information on AI mechanisms such as machine learning (ML), deep learning (DL), and convolutional neural networks (CNN) as they are being incorporated into pre-operative, intraoperative, and post-operative stages of cataract surgery. AI is currently integrated in the pre-operative stage of cataract surgery to calculate intraocular lens (IOL) power and diagnose cataracts with slit-lamp microscopy and retinal imaging. During the intraoperative stage, AI has been applied to risk calculation, tracking surgical workflow, multimodal imaging data analysis, and instrument location via the use of "smart instruments". AI is also involved in predicting post-operative complications, such as posterior capsular opacification and intraocular lens dislocation, and organizing follow-up patient care. Challenges such as limited imaging dataset availability, unstandardized deep learning analysis metrics, and lack of generalizability to novel datasets currently present obstacles to the enhanced application of AI in cataract surgery. Upon addressing these barriers in upcoming research, AI stands to improve cataract screening accessibility, junior physician training, and identification of surgical complications through future applications of AI in cataract surgery. Competing Interests: The authors report no conflicts of interest in this work. (© 2024 Ahuja et al.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu