目的:探究白内障患者人工晶状体(IOL)植入后Kappa角变化趋势及与IOL稳定性的关系。方法:选择2023-02/2024-01于我院接受IOL植入的白内障患者120例120眼为研究对象,其中左眼56眼,右眼64眼。比较手术前后眼科检查结果、Kappa角的分布,以...目的:探究白内障患者人工晶状体(IOL)植入后Kappa角变化趋势及与IOL稳定性的关系。方法:选择2023-02/2024-01于我院接受IOL植入的白内障患者120例120眼为研究对象,其中左眼56眼,右眼64眼。比较手术前后眼科检查结果、Kappa角的分布,以及术后不同时点Kappa角变化和IOL稳定性。Pearson相关及多重线性回归分析IOL旋转稳定性与各眼科检查指标的相关性。多元线性回归分析Kappa角与IOL稳定性的相关性并绘制变化趋势图。结果:术后Kappa角逐渐减小且减小程度逐渐降低(均P<0.05),IOL旋转度数也逐渐减小(P<0.05)。术后2 mo IOL旋转度数与术后1 d AL、LT、K1、K2、PD、Kappa角呈正相关(均P<0.05),与术后1 d ACD呈负相关(均P<0.05)。术后1 d AL、ACD、PD、Kappa角对IOL旋转度数产生显著影响(均P<0.05)。随着Kappa角增大,IOL旋转度数逐渐增大,即IOL稳定性逐渐降低(P<0.05)。结论:白内障患者IOL植入后Kappa角显著减小,前期减小幅度较大,而后趋于稳定,IOL稳定性随Kappa角的增大而降低。展开更多
Background and Objective:Refractive surgery has evolved significantly,with artificial intelligence(AI)offering new possibilities for enhancing patient selection,surgical planning,and postoperative outcome prediction.W...Background and Objective:Refractive surgery has evolved significantly,with artificial intelligence(AI)offering new possibilities for enhancing patient selection,surgical planning,and postoperative outcome prediction.While AI has demonstrated promising applications,its integration into clinical practice remains inconsistent due to challenges in data standardization,model interpretability,and regulatory concerns.This review examines the current applications,limitations,and future directions of AI in refractive surgery,with a focus on its role in laser vision correction(LVC)and phakic intraocular lens(IOL)implantation.Methods:A literature review was conducted using peer-reviewed studies published between January 2010 and October 2024,sourced from databases including Google Scholar,PubMed,Embase,and Web of Science.Studies were selected based on predefined inclusion criteria,covering AI applications in refractive surgery.A total of 33 key studies(16 on LVC and 17 on phakic IOLs)were analyzed,focusing on machine learning and deep learning techniques used for patient selection,surgical optimization,and complication prediction.Only English-language studies were included.Key Content and Findings:AI models utilizing structured tabular data,imaging,and multimodal inputs have demonstrated superior performance in predicting surgical outcomes and refining patient selection compared to traditional methods.Machine learning approaches such as random forests,extreme gradient boosting,and ensemble techniques,alongside deep learning architectures like convolutional neural networks and generative models,have improved risk assessment and surgical planning.In LVC,AI enhances ectasia risk assessment,keratoconus detection,and myopic regression prediction.In phakic IOL implantation,AI improves postoperative vault prediction,lens sizing,and refractive error estimation.Multimodal AI systems integrating imaging,textual data,and clinical parameters hold promise for more comprehensive patient evaluations.However,challenges such as data heterogeneity,limited external validation,and regulatory barriers hinder widespread clinical adoption.Conclusions:AI is transforming refractive surgery by enhancing precision,personalization,and patient safety.Its integration into clinical workflows has the potential to improve surgical outcomes and patient satisfaction.Future efforts should focus on advancing multimodal AI,improving model generalizability,and addressing ethical and regulatory challenges to fully harness AI’s potential in refractive surgery.展开更多
文摘目的:探究白内障患者人工晶状体(IOL)植入后Kappa角变化趋势及与IOL稳定性的关系。方法:选择2023-02/2024-01于我院接受IOL植入的白内障患者120例120眼为研究对象,其中左眼56眼,右眼64眼。比较手术前后眼科检查结果、Kappa角的分布,以及术后不同时点Kappa角变化和IOL稳定性。Pearson相关及多重线性回归分析IOL旋转稳定性与各眼科检查指标的相关性。多元线性回归分析Kappa角与IOL稳定性的相关性并绘制变化趋势图。结果:术后Kappa角逐渐减小且减小程度逐渐降低(均P<0.05),IOL旋转度数也逐渐减小(P<0.05)。术后2 mo IOL旋转度数与术后1 d AL、LT、K1、K2、PD、Kappa角呈正相关(均P<0.05),与术后1 d ACD呈负相关(均P<0.05)。术后1 d AL、ACD、PD、Kappa角对IOL旋转度数产生显著影响(均P<0.05)。随着Kappa角增大,IOL旋转度数逐渐增大,即IOL稳定性逐渐降低(P<0.05)。结论:白内障患者IOL植入后Kappa角显著减小,前期减小幅度较大,而后趋于稳定,IOL稳定性随Kappa角的增大而降低。
文摘Background and Objective:Refractive surgery has evolved significantly,with artificial intelligence(AI)offering new possibilities for enhancing patient selection,surgical planning,and postoperative outcome prediction.While AI has demonstrated promising applications,its integration into clinical practice remains inconsistent due to challenges in data standardization,model interpretability,and regulatory concerns.This review examines the current applications,limitations,and future directions of AI in refractive surgery,with a focus on its role in laser vision correction(LVC)and phakic intraocular lens(IOL)implantation.Methods:A literature review was conducted using peer-reviewed studies published between January 2010 and October 2024,sourced from databases including Google Scholar,PubMed,Embase,and Web of Science.Studies were selected based on predefined inclusion criteria,covering AI applications in refractive surgery.A total of 33 key studies(16 on LVC and 17 on phakic IOLs)were analyzed,focusing on machine learning and deep learning techniques used for patient selection,surgical optimization,and complication prediction.Only English-language studies were included.Key Content and Findings:AI models utilizing structured tabular data,imaging,and multimodal inputs have demonstrated superior performance in predicting surgical outcomes and refining patient selection compared to traditional methods.Machine learning approaches such as random forests,extreme gradient boosting,and ensemble techniques,alongside deep learning architectures like convolutional neural networks and generative models,have improved risk assessment and surgical planning.In LVC,AI enhances ectasia risk assessment,keratoconus detection,and myopic regression prediction.In phakic IOL implantation,AI improves postoperative vault prediction,lens sizing,and refractive error estimation.Multimodal AI systems integrating imaging,textual data,and clinical parameters hold promise for more comprehensive patient evaluations.However,challenges such as data heterogeneity,limited external validation,and regulatory barriers hinder widespread clinical adoption.Conclusions:AI is transforming refractive surgery by enhancing precision,personalization,and patient safety.Its integration into clinical workflows has the potential to improve surgical outcomes and patient satisfaction.Future efforts should focus on advancing multimodal AI,improving model generalizability,and addressing ethical and regulatory challenges to fully harness AI’s potential in refractive surgery.
