基于人工智能技术的斜视诊疗进展
Progress in diagnosis and treatment of strabismus based on artificial intelligence technology
郭勇麟 1陈墨馨 1刘哲源 2李奕霏 1王子琦 1舒琴 1李琳1
作者信息
- 1. 上海交通大学医学院附属第九人民医院眼科,上海 200011;上海市眼眶病眼肿瘤重点实验室,上海 200011
- 2. 上海交通大学电子信息与电气工程学院,上海 200240
- 折叠
摘要
斜视是由于中枢管制失调、眼外肌力量失衡,导致双眼不能同时注视目标的一种眼病,通常在儿童期发病,可造成弱视、视功能损害、斜颈等发育不良和心理障碍.该病对患者、家庭乃至社会都会产生消极影响.早诊断、早治疗可避免造成视力和立体视的永久损害.然而,目前斜视的诊断高度依赖眼科医师专业检查,效率低、覆盖面小,而常规的入校筛查无法反映斜视度,且准确率低.因此如何提高斜视筛检效率是具有重要意义的热点问题.该文从人工智能技术发展的角度,基于近年来国内外的斜视诊疗现状,综述斜视领域中人工智能模型和算法的研究进展及不足,为进一步探索人工智能辅助的斜视诊疗前景提供参考.
Abstract
Strabismus,misalignment of the eyes arising from central nervous system dysregulation and extraocular muscles imbalance,commonly manifests in childhood,leading to amblyopia,binocular vision dysfunction,torticollis and other developmental and psychological disorders.This exerts a negative impact on individuals,families and society.Timely diagnosis and intervention are crucial to prevent permanent damage to vision and stereopsis.Presently,strabismus diagnosis is reliant on the ophthalmologists' evaluations which results in a lack of efficiency and coverage.However,routine school screening proves inadequate in assessing strabismus degree with low accuracy.Therefore,how to improve the efficiency of strabismus screening is an issue of great importance.This paper delves into the present landscape of strabismus diagnosis and treatment,considering both local and global research advancements.It focuses on the evolution of artificial intelligence technology,illuminating the utilization of artificial intelligence models and algorithms in strabismus.By pinpointing and exploring their strengths and limitations,it offers valuable insights,paving the way for future investigations into artificial intelligence-assisted strabismus diagnosis and treatment.
关键词
斜视/人工智能/诊断/治疗/算法Key words
strabismus/artificial intelligence/diagnosis/treatment/algorithm引用本文复制引用
基金项目
上海市科技创新行动计划生物医药科技支撑专项(23S31900500)
上海交通大学中国医院发展研究院定向委托项目(CHDI-2022-DX-02)
上海市高水平地方高校创新团队项目(SSMU-ZDCX20180401)
上海交通大学医学院大学生创新性训练计划(1824026)
出版年
2024
NETL
NSTL
维普
万方数据