国际眼科纵览 ›› 2025, Vol. 49 ›› Issue (2): 95-100.doi: 10.3760/cma.j.cn115500-20241103-25203

• • 上一篇    下一篇

基于人工智能的IOL屈光度计算公式

张然  王震宇  杨一佺  王进达   

  1. 首都医科大学附属北京同仁医院 北京同仁眼科中心 眼科学与视觉科学北京市重点实验室,北京 100730
  • 收稿日期:2024-11-03 出版日期:2025-04-22 发布日期:2025-04-10
  • 通讯作者: 王进达,Email:wjd3636@163.com
  • 基金资助:
    北京市自然科学基金(7242038);北京市科技新星计划(20220484224)

Intraocular lens power calculation formulas based on artificial intelligence

Zhang Ran, Wang Zhenyu, Yang Yiquan, Wang Jinda   

  1. Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University; Beijing Key Laboratory of Ophthalmology and Visual Sciences, Beijing 100730, China
  • Received:2024-11-03 Online:2025-04-22 Published:2025-04-10
  • Contact: Wang Jinda, Email: wjd3636@163.com
  • Supported by:
     Beijing Natural Science Foundation (7242038); Beijing Nova Program (20220484224)

摘要: 目前临床上已有多种IOL屈光度计算公式,但在预测准确性方面仍有提升空间,尤其是在极端眼轴长度或复杂眼部条件的患者中表现不足。为提高IOL屈光度计算的准确性,基于人工智能(artificial intelligence,AI)的IOL屈光度计算公式被人们开发并寄予厚望。本文总结了近几年新开发的基于AI的IOL屈光度计算公式,包括纯数据驱动的IOL屈光度计算公式(Hill-RBF公式、Karmona公式、Nallasamy公式等),光学理论结合AI的IOL屈光度计算公式(Kane公式、PEARL-DGS公式、LISA-PPV公式等),其他公式(Hoffer QST公式、Ladas公式、Zhu-Lu公式等),这些公式在处理极端眼轴和复杂眼部条件时展现出更高的预测精度,为临床医师提供更多的公式选择,进一步提高IOL屈光度的预测准确性。 (国际眼科纵览,2025, 49:95-100

关键词: 人工智能, IOL, 计算公式

Abstract: Currently, there are various intraocular lens (IOL) power calculation formulas available in clinical practice, but there is still room for improvement in their prediction accuracy, especially in cases of extreme axial lengths and complex ocular conditions. To enhance the accuracy of IOL power calculations, IOL power calculation formulas based on artificial intelligence (AI) have been developed and hold great promise. This article summarizes the newly developed AI-based IOL power calculation formulas in recent years, including pure data-driven IOL power calculation formulas (Hill RBF formula, Karmana formula, Nallasamy formula, etc.), optical theory combined with AI IOL power calculation formulas (Kane formula, PEARL-DGS formula, LISA-PPV formula, etc.), and other formulas (Hoffer QST formula, Ladas formula, Zhu Lu formula, etc.). These formulas demonstrate higher prediction accuracy when dealing with extreme eye axis and complex eye conditions, aiming to provide clinicians with more options and further improve the prediction accuracy of IOL power. (Int Rev Ophthalmol, 2025, 49:  95-100)

Key words: artificial intelligence, IOL, formula