Stopyra Wiktor, Grzybowski Andrzej
MW-Med Eye Centre, 31-416 Krakow, Poland.
Medical Institute, University of Applied Sciences, 34-400 Nowy Targ, Poland.
J Clin Med. 2025 Sep 18;14(18):6585. doi: 10.3390/jcm14186585.
The accurate intraocular lens (IOL) power calculation is essential in phacoemulsification. The latest IOL power calculation formulas, and their new classification and method to assess their accuracy, were described and analyzed. Narrative review. The manuscript includes articles on IOL power calculation published between 2019 and 2025. The following key words, such as "phacoemulsification", "IOL power calculation formula", "AI-based formulas", "IOL power selection", "IOL formulas classification", IOL prediction" were used to identify papers by searching medical databases (Pubmed/MEDLINE, Google Scholar, Crossref). PRISMA methodology was used to select articles. Finally, 33 peer-reviewed English-language studies with a sample size of at least 120 eyes were included in the analysis. Ten IOL power calculation formulas that have been introduced and published over the past 5 years were included in the study. Five of them are artificial intelligence based (Karmona, Hoffer QST, Nallasamy, Zhu-Lu and Zeiss-AI), four are vergence (Emmetropia Verifying Optical, Naeser 2, Voytsekhivskyy Regression Function-Gender and Castrop), and one is ray-tracing (the O formula). In this review, the formulas are introduced and analyzed, with a discussion of selected studies assessing the accuracy of these IOL power calculation methods. New IOL power calculation formulas are constantly being developed. They are mostly based on artificial intelligence. New methods are still being sought to assess the accuracy of formulas, and root mean square absolute error is one of them.
准确的人工晶状体（IOL）屈光力计算在白内障超声乳化手术中至关重要。本文描述并分析了最新的IOL屈光力计算公式及其新的分类和评估其准确性的方法。叙述性综述。该手稿纳入了2019年至2025年间发表的关于IOL屈光力计算的文章。通过检索医学数据库（PubMed/MEDLINE、谷歌学术、CrossRef），使用了“白内障超声乳化术”、“IOL屈光力计算公式”、“基于人工智能的公式”、“IOL屈光力选择”、“IOL公式分类”、“IOL预测”等关键词来识别论文。采用PRISMA方法选择文章。最后，分析纳入了33篇样本量至少为120只眼的同行评审英文研究。本研究纳入了过去5年中引入并发表的10种IOL屈光力计算公式。其中5种基于人工智能（卡尔莫纳、霍弗QST、纳拉萨米、朱-陆和蔡司-人工智能），4种基于屈光（正视化验证光学、内泽尔2、沃伊采希夫斯基回归函数-性别和卡斯特罗普），1种基于光线追踪（O公式）。在本综述中，介绍并分析了这些公式，并讨论了评估这些IOL屈光力计算方法准确性的选定研究。新的IOL屈光力计算公式不断涌现。它们大多基于人工智能。仍在寻求评估公式准确性的新方法，均方根绝对误差就是其中之一。
