Research progress and application prospects of DeepSeek in ophthalmology
10.3760/cma.j.cn115989-20250607-00184
- VernacularTitle:DeepSeek在眼科的研究进展和应用前景
- Author:
Shiyang NIU
1
;
Lijun QU
1
Author Information
1. 哈尔滨医科大学附属第二医院眼科，哈尔滨 150001
- Publication Type:Journal Article
- Keywords:
DeepSeek;
Artificial intelligence;
Ophthalmology;
Deep learning;
Large language models
- From:
Chinese Journal of Experimental Ophthalmology
2025;43(11):1046-1052
- CountryChina
- Language:Chinese
-
Abstract:
Ophthalmology is undergoing a profound transformation driven by artificial intelligence (AI), represented by deep learning technologies.As the frontier of AI advancement, large language models (LLMs) demonstrate substantial potential in processing massive unstructured medical text and assisting with complex clinical reasoning.The DeepSeek series models, as exemplary advanced LLMs, provide novel opportunities for ophthalmic clinical practice, scientific research, and administrative management through their exceptional reasoning capabilities, remarkable cost-effectiveness, and open-source nature.Given research specifically targeting DeepSeek's ophthalmological applications remains nascent, this review systematically delineates its potential in diagnostic assistance, natural language processing, and functioning as the linguistic core for multimodal systems, which is achieved by precisely defining its LLM capabilities and benchmarking against validated applications of comparable advanced models.This article thoroughly discusses critical challenges including model " hallucinations", data governance, and regulatory approvals in clinical practice, aiming to provide early, rigorous academic guidance for this emerging interdisciplinary field and promote the healthy, standardized, and innovative advancement of DeepSeek technology in ophthalmology.
