人工智能(artificial intelligence，AI)为解决中国患者“看病难”问题提供了可行方案。眼科AI已实现为患者提供筛查、远程诊断及治疗建议等方面的服务，能显著减轻医疗资源不足的压力和患者的经济负担。而AI的应用过程中，给医疗管理带来的挑战应引起重视。本文从医疗管理的角度，总结分析AI在眼科医疗过程中，尤其是交接环节中出现的主要问题，提出对策与建议，并讨论AI在眼科医疗的应用展望。

    Artificial intelligence (AI) has been proposed as a potential solution to address the shortage of ophthalmologists in China. With the increasingly extensive application of AI in the field of ophthalmology, many potential patients with eye diseases have access to a higher quality of medical services. At the same time, new challenges will emerge and proliferate with the advancement of AI application. This paper focuses on the patient handoffs process and discusses two challenges brought by the application of AI, namely “communication” and “standardization”. Natural language processing techniques and the development of standardized databases are proposed to solve each of these challenges. The application prospects of AI in ophthalmology are eventually discussed.

泪器疾病是一类常见的眼科疾病，其诊疗过程复杂，治疗方法精细，涉及多种临床数据及影像资料。现有研究表明，随着人工智能（artificial intelligence，AI）技术，尤其是机器学习和深度学习的发展，AI在泪器疾病的早期筛查、精确诊断和个性化治疗中展现了巨大的应用潜力。AI能够通过高效的图像分析、多模态数据融合及深度学习算法，提供更加精确的疾病识别和治疗方案，并且能够对患者的病情进行定期监测和动态调整，提升治疗效果。然而，其仍面临诸多挑战，如多模态数据融合的复杂性、模型泛化能力的局限以及实时预测和动态调整的需求等，需要通过持续的技术创新、算法优化和跨学科合作来实现。文章对当前AI在泪器疾病诊疗中的应用现状进行了全面梳理和总结，深入分析了AI技术在诊断与治疗中的优势与局限，特别强调了AI与新兴技术的结合在优化临床决策支持系统方面的重要性。通过分析现有的挑战与技术融合策略，文章提出了AI在泪器疾病诊疗中的发展方向，旨在为未来的研究者提供创新性的思路，为眼科领域的临床实践提供有价值的参考，助力泪器疾病的精准医疗和个性化治疗的发展。

Lacrimal disorders are common ophthalmic conditions characterized by complex diagnostic and treatment processes, involving intricate therapeutic approaches and diverse clinical and imaging data. Recent studies have indicated that with the advancements in artificial intelligence (AI) technologies, particularly in machine learning and deep learning, AI demonstrates significant potential in the early screening, accurate diagnosis, and personalized treatment of lacrimal disorders. AI has the ability to provide more precise disease identification and treatment strategies through efficient image analysis, multimodal data fusion, and deep learning algorithms. Additionally, it enables regular monitoring and dynamic adjustment of patients' conditions, improving treatment outcomes. However, several challenges persist, such as the complexity of multimodal data integration, limitations in model generalization capabilities, and the need for real-time prediction and dynamic adjustments, all of which necessitate continuous technological innovations, algorithm optimization, and interdisciplinary collaborations. This paper provides a comprehensive review of the current status of AI applications in the diagnosis and treatment of lacrimal disorders, analyzing the advantages and limitations of AI in clinical practice. It especially emphasizes the importance of integrating AI with emerging technologies to optimize clinical decision support systems. By addressing the existing challenges and exploring strategies for technological integration, this paper proposes future directions for the development of AI in lacrimal disorder diagnosis and treatment, aiming to offer innovative perspectives for future researchers and valuable references for clinical practice in the field of ophthalmology, ultimately contributing to the advancement of precision medicine and personalized treatment for lacrimal disorders.

近年来随着人类生活方式的改变、用眼频率的增加，眼科药物的市场需求持续增长，但是目前眼病治疗仍面临“缺医少药”的困境。由于新药研发面临成本高、周期长、成功率低的风险，眼科药物创新迭代的进程日趋缓慢。人工智能(artificial intelligence，AI)作为一种全新的技术手段，有望赋能眼科药物研发的全过程，包括药物靶点发现、化合物筛选、药物动力学模型创新与临床试验开展等，以期为眼科药物研发“降本增效”。且随着大数据体系的完善、硬件计算力的提升以及生命科学与智能科学的深度融合，AI在眼科药物研发中的作用将进一步得到提升，助力眼科药物研发实现从精准化到智能化的跨越。

With the change of human lifestyle and overuse of eyes in recent years, the market demand for ophthalmic drugs continues to grow. However, the ocular therapy is still facing the shortage of doctors and drugs. Due to the risk of high cost, long lead time and low success rate, the process of novel ophthalmic drug innovation and iteration is getting slower. As an emerging technology, artificial intelligence is expected to enable the whole process of ophthalmic drug discovery and development, including drug target discovery, compound screening, pharmacokinetic model innovation and clinical trials, thus reducing R&D costs and increase efficiency for ophthalmic drug discovery and development. In addition, with the improvement of big data, hardware calculation and the deep integration of life science and intelligent science, the role of artificial intelligence in ophthalmic drug discovery and development will be significant improved , contributing to achieve the leap from precision to intelligence.

手术前常规检查在临床诊疗中被广泛应用，但在一些低风险择期手术前对患者进行常规检查，对提高医疗质量并无帮助，反而降低了医疗效率，增加了医疗费用。为提高效率，一些地区、机构和专家学者陆续通过宣传教育、发表共识、制定指南等方式控制无指征术前常规检查，但效果仍依赖于执业者的重视程度和专业水平。大数据机器学习方法以其标准化、自动化的特点为解决这一问题提供了新的思路。在回顾已有研究的基础上，我们抽取2017至2019年在中山大学中山眼科中心进行眼科手术的3.4万名患者的病史和体格检查资料大数据，涵盖年龄、性别等口学信息，诊断、既往疾病等病史信息，视功能、入院时身体质量指数(BMI)等体格检查信息。并以此为基础使用机器学习方法预测术前胸部X线检查是否存在异常，受试者操作特性曲线(receiver operating characteristic curve，ROC)曲线下面积达到0.864，预测准确率可达到81.2%，对大数据机器学习精简术前常规检查的新方式进行了先期探索。

Preoperative routine tests are widely prescribed in clinical settings. However, these tests do not help improving the quality of medical care in low-risk elective surgery. Instead, they are associated with lower efficiency and increasing fees. To improve the efficiency, many regions, institutions, and scholars have attempted to reduce preoperative routine tests without indications through propaganda, education, consensus, and guidelines. Nevertheless, the effects are still highly dependent on the expertise and emphasis of practitioners. Machine learning based on big data provide a new solution with its standardization and automation. Through literature review, we extracted the big data, including demographic features such as sex and age, histories including diagnosis and chronic diseases, and physical examination features such as visual function and body mass index. A total of 34 000 patients undergone ocular surgeries in Zhongshan Ophthalmic Center, Sun Yat-sen university from 2017 to 2019. Machine learning was adopted to predict the risk of finding abnormalities in chest X-ray examination, with an accuracy of 81.2%. Area under the Receiver Operating Characteristic curve was 0.864. The study could be an early exploration into the field of simplifying preoperative tests by machine learning.