[关键词]
[摘要]
随着白内障手术进入屈光时代，术后残留屈光不正成为影响视觉质量的关键因素，而人工晶状体(IOL)屈光度计算准确性受眼部生物参数、计算公式及晶状体常数等多重影响，目前临床广泛使用的晶状体常数多基于西方人群，与我国人群短眼轴、陡角膜等眼部特征存在适配偏差，因此常数的个性化优化成为研究热点。文章主要阐述晶状体常数优化在提高IOL屈光度计算准确性的研究进展，目前晶状体常数优化方式尚未形成共识，单常数公式可通过迭代法优化，多常数公式需结合线性或非线性策略，简化梯度下降法、数据驱动技术等为优化提供了新路径，但需交叉验证评估性能； 基于眼轴长度、角膜曲率、前房深度的分段优化在特殊解剖结构患者中效果显著，可有效降低不同分组人群的术后屈光误差，但部分极端病例仍存在局限性； 多参数交互作用对复杂病例的预测准确性影响显著，新一代整合多参数或引入AI算法的公式虽有提升，但常数优化仍具有价值。当前，多参数复杂关系、术中动态变化等问题仍需深入研究，未来需开展针对性人群优化研究、开发实时监测技术及创新 IOL 设计，以进一步趋近术后屈光零误差目标。
[Key word]
[Abstract]
With the advent of the refractive era of cataract surgery, postoperative residual refractive error has become a key factor affecting visual quality. The accuracy of intraocular lens(IOL)power calculation is affected by multiple factors, including ocular biological parameters, calculation formulas, and lens constants. Currently, the lens constants widely used in clinical practice are mostly based on Western populations, which have a mismatch with the ocular characteristics of the Chinese population, such as shorter axial length and steeper cornea. Therefore, the personalized optimization of the constant has become a research hotspot. This review primarily summarizes the research progress on lens constant optimization in improving the accuracy of IOL power calculation. Currently, there is no consensus on lens constant optimization methods. Single-constant formulas can be optimized through iterative methods, while multi-constant formulas require the combination of linear or nonlinear strategies. Simplified gradient descent and data-driven techniques offer new optimization pathways, but cross-validation is needed to evaluate their performance. Segmented optimization based on axial length, corneal curvature, and anterior chamber depth has shown significant effectiveness in patients with special anatomical structures, effectively reducing postoperative refractive errors in different patient groups, but limitations remain in some extreme cases. The interaction of multiple parameters significantly impacts the predictive accuracy of complex cases. While new-generation formulas integrating multiple parameters or incorporating AI algorithms have improved accuracy, constant optimization still holds value. Currently, the complex relationships between multiple parameters and intraoperative dynamic changes require further in-depth research. Future research, including targeted population optimization studies, the development of real-time monitoring technologies, and innovative IOL designs, may make achieving zero postoperative refractive error a possibility.
[中图分类号]
[基金项目]
贵州省卫生健康委科学技术基金项目(No.gzwkj2023-435)
