Chen Bichi, Tian Li, Tian Fuyue, Yang Qiaochu, Ruan Ying, Li Ying, Cao Min, Wu Chuanyan, Yang Maoyuan, Xu Suzhong, Deng Ruzhi
Vision X Medical Technology Co., Ltd., Shanghai, China.
National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China.
Front Cell Dev Biol. 2025 May 22;13:1608494. doi: 10.3389/fcell.2025.1608494. eCollection 2025.
To develop and validate machine learning (ML) models for predicting cycloplegic spherical equivalent refraction (SER) using non-cycloplegic parameters, addressing challenges in pediatric ophthalmic assessments.
A prospective cohort of 2,274 Chinese children (4,548 eyes) aged 3∼16 years was stratified into development ( = 1819) and validation ( = 455) datasets. Six ML models (linear regression, random forest, extreme gradient boosting, multilayer perceptron, support vector machine, and light gradient boosting machine) were trained on demographics, non-cycloplegic refractive error, and ocular biometrics. Model performance was evaluated using , mean error (ME), mean absolute error (MAE), and clinical accuracy (proportions within ±0.50 D/±1.00 D).
In the validation dataset, ML models predicted cycloplegic SER with high (0.920∼0.934), low ME (-0.004∼0.015 D) and MAE (0.385∼0.413 D). The multilayer perceptron model achieved the highest accuracy ( = 0.934, MAE = 0.385 D), with 73.08% and 94.29% of predictions within ±0.50 D and ±1.00 D, respectively. Performance was optimal in children aged 7∼10 years (77.17∼79.70% within ±0.50 D) and those with low myopia (-3.00 to -0.50 D; 83.09∼83.56% within ±0.50 D). Non-cycloplegic measurements systematically overestimated myopia (mean difference: -0.39 ± 0.71 D, < 0.001), particularly in younger children and hyperopic eyes.
ML models provide accurate estimates of cycloplegic SER using non-cycloplegic parameters, offering a practical alternative for pediatric refractive assessments when cycloplegia is infeasible.
开发并验证用于使用非散瞳参数预测散瞳等效球镜度（SER）的机器学习（ML）模型，以应对儿科眼科评估中的挑战。
将2274名年龄在3至16岁的中国儿童（4548只眼）的前瞻性队列分为开发数据集（n = 1819）和验证数据集（n = 455）。基于人口统计学、非散瞳屈光不正和眼部生物特征训练了六种ML模型（线性回归、随机森林、极端梯度提升、多层感知器、支持向量机和轻梯度提升机）。使用决定系数（R²）、平均误差（ME）、平均绝对误差（MAE）和临床准确性（±0.50 D/±1.00 D范围内的比例）评估模型性能。
在验证数据集中，ML模型预测散瞳SER的R²较高（0.920至0.934），ME较低（-0.004至0.015 D），MAE较低（0.385至0.413 D）。多层感知器模型实现了最高准确性（R² = 0.934，MAE = 0.385 D），分别有73.08%和94.29%的预测在±0.50 D和±1.00 D范围内。在7至10岁儿童（±0.50 D范围内为77.17%至79.70%）和低度近视儿童（-3.00至-0.50 D；±0.50 D范围内为83.09%至83.56%）中性能最佳。非散瞳测量系统性高估了近视（平均差异：-0.39 ± 0.71 D，P < 0.001），尤其是在年幼儿童和远视眼中。
ML模型使用非散瞳参数提供了准确的散瞳SER估计值，当散瞳不可行时，为儿科屈光评估提供了一种实用的替代方法。
