Enhancing Early Keratoconus Detection With Multimodal Machine Learning: Integrating Tomography, Biomechanics, and Clinical Risk Factors

IF 4.2 1区医学 Q1 OPHTHALMOLOGY

American Journal of Ophthalmology Pub Date : 2025-08-19 DOI:10.1016/j.ajo.2025.08.035

Kaiyue Du , Rongmei Peng , Yueguo Chen , Bowei Yuan , Haoran Wu , Tiehong Chen , Jianing Zhu , Xunshan Zu , Jiaojiao Wang , Jing Cui , Liang Han , Jing Hong

{"title":"Enhancing Early Keratoconus Detection With Multimodal Machine Learning: Integrating Tomography, Biomechanics, and Clinical Risk Factors","authors":"Kaiyue Du , Rongmei Peng , Yueguo Chen , Bowei Yuan , Haoran Wu , Tiehong Chen , Jianing Zhu , Xunshan Zu , Jiaojiao Wang , Jing Cui , Liang Han , Jing Hong","doi":"10.1016/j.ajo.2025.08.035","DOIUrl":null,"url":null,"abstract":"<div><h3>OBJECTIVE</h3><div>To develop and validate a machine learning (ML) diagnostic system that integrates Scheimpflug tomography, corneal biomechanics, and clinical risk factors (CRF) to enhance the early detection of keratoconus (KC).</div></div><div><h3>DESIGN</h3><div>Prospective, multicenter, cross-sectional study.</div></div><div><h3>PARTICIPANTS</h3><div>Patients diagnosed with KC and individuals evaluated in preoperative refractive surgery clinics.</div></div><div><h3>METHODS</h3><div>Demographic, lifestyle, and clinical ophthalmic data, including Pentacam and Corvis ST measurements, were collected from patients with KC and refractive surgery candidates across 5 centers between 2020 and 2024. The dataset was divided into training, validation, internal test, and external test sets. Least absolute shrinkage and selection operator regression was used to identify predictive variables. Six ML models were trained using 4 feature sets: CRF, device-derived parameters, combined features, and selected features.</div></div><div><h3>MAIN OUTCOME MEASURES</h3><div>Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC).</div></div><div><h3>RESULTS</h3><div>The dataset included 1035 eyes from 1035 participants across 5 centers: 590 normal controls, 157 eyes with forme fruste keratoconus (FFKC), 143 with subclinical KC, and 145 with clinical KC. For FFKC detection, the post-feature selection CatBoost model achieved the highest accuracy (AUROC = 0.975), outperforming the combined-feature (AUROC = 0.963), CRF-only (AUROC = 0.856), and device-only models (AUROC = 0.885) in the validation set. This model also outperformed the tomographic and biomechanical index in internal (AUROC = 0.976 vs 0.813; <em>P</em> = .048) and external testing (AUROC = 0.952 vs 0.847; <em>P</em> = .012). For subclinical and clinical KC, external testing yielded near-perfect performance (AUROC = 0.991 and 1.000, respectively).</div></div><div><h3>CONCLUSIONS</h3><div>A multimodal ML system integrating CRF, tomography, and biomechanics improved early KC detection, particularly for FFKC. This approach may enhance clinical decision-making and screening for refractive surgery candidates.</div></div>","PeriodicalId":7568,"journal":{"name":"American Journal of Ophthalmology","volume":"280 ","pages":"Pages 334-346"},"PeriodicalIF":4.2000,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"American Journal of Ophthalmology","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0002939425004404","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"OPHTHALMOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

OBJECTIVE

To develop and validate a machine learning (ML) diagnostic system that integrates Scheimpflug tomography, corneal biomechanics, and clinical risk factors (CRF) to enhance the early detection of keratoconus (KC).

DESIGN

Prospective, multicenter, cross-sectional study.

PARTICIPANTS

Patients diagnosed with KC and individuals evaluated in preoperative refractive surgery clinics.

METHODS

Demographic, lifestyle, and clinical ophthalmic data, including Pentacam and Corvis ST measurements, were collected from patients with KC and refractive surgery candidates across 5 centers between 2020 and 2024. The dataset was divided into training, validation, internal test, and external test sets. Least absolute shrinkage and selection operator regression was used to identify predictive variables. Six ML models were trained using 4 feature sets: CRF, device-derived parameters, combined features, and selected features.

MAIN OUTCOME MEASURES

Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC).

RESULTS

The dataset included 1035 eyes from 1035 participants across 5 centers: 590 normal controls, 157 eyes with forme fruste keratoconus (FFKC), 143 with subclinical KC, and 145 with clinical KC. For FFKC detection, the post-feature selection CatBoost model achieved the highest accuracy (AUROC = 0.975), outperforming the combined-feature (AUROC = 0.963), CRF-only (AUROC = 0.856), and device-only models (AUROC = 0.885) in the validation set. This model also outperformed the tomographic and biomechanical index in internal (AUROC = 0.976 vs 0.813; P = .048) and external testing (AUROC = 0.952 vs 0.847; P = .012). For subclinical and clinical KC, external testing yielded near-perfect performance (AUROC = 0.991 and 1.000, respectively).

CONCLUSIONS

A multimodal ML system integrating CRF, tomography, and biomechanics improved early KC detection, particularly for FFKC. This approach may enhance clinical decision-making and screening for refractive surgery candidates.

查看原文本刊更多论文

利用多模态机器学习增强圆锥角膜的早期检测：整合断层扫描、生物力学和临床危险因素。

目的：开发并验证一种结合Scheimpflug断层扫描、角膜生物力学和临床危险因素（CRF）的机器学习（ML）诊断系统，以提高圆锥角膜（KC）的早期发现。设计：前瞻性、多中心、横断面研究。参与者：被诊断为KC的患者和在术前屈光手术诊所评估的个体。方法：从2020年至2024年五个中心的KC患者和屈光手术候选人中收集人口统计学、生活方式和临床眼科数据，包括Pentacam和Corvis ST测量值。数据集分为训练集、验证集、内部测试集和外部测试集。最小绝对收缩和选择算子回归用于识别预测变量。使用四个特征集训练六个ML模型：CRF，设备衍生参数，组合特征和选择特征。主要结果测量：使用受试者工作特征曲线下面积（AUROC）评估模型性能。结果：数据集包括来自五个中心的1,035名参与者的1,035只眼睛：590名正常对照，157只形成锥状角膜（FFKC）， 143只亚临床KC和145只临床KC。对于FFKC检测，后特征选择CatBoost模型的准确率最高（AUROC = 0.975），优于组合特征（AUROC = 0.963），仅crf （AUROC = 0.856）和仅设备模型（AUROC = 0.885）。该模型在内部验证（AUROC = 0.976 vs. 0.813; p = 0.048）和外部验证（AUROC = 0.952 vs. 0.847; p = 0.012）中也优于断层扫描和生物力学指标。对于亚临床和临床KC，外部验证产生了近乎完美的性能（AUROC分别为 = 0.991和1.000）。结论：结合CRF、断层扫描和生物力学的多模态ML系统改善了早期KC的检测，特别是FFKC。这种方法可以提高屈光手术候选人的临床决策和筛选。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

American Journal of Ophthalmology 医学-眼科学

CiteScore

9.20

自引率

7.10%

发文量

406

审稿时长

36 days

期刊介绍： The American Journal of Ophthalmology is a peer-reviewed, scientific publication that welcomes the submission of original, previously unpublished manuscripts directed to ophthalmologists and visual science specialists describing clinical investigations, clinical observations, and clinically relevant laboratory investigations. Published monthly since 1884, the full text of the American Journal of Ophthalmology and supplementary material are also presented online at www.AJO.com and on ScienceDirect. The American Journal of Ophthalmology publishes Full-Length Articles, Perspectives, Editorials, Correspondences, Books Reports and Announcements. Brief Reports and Case Reports are no longer published. We recommend submitting Brief Reports and Case Reports to our companion publication, the American Journal of Ophthalmology Case Reports. Manuscripts are accepted with the understanding that they have not been and will not be published elsewhere substantially in any format, and that there are no ethical problems with the content or data collection. Authors may be requested to produce the data upon which the manuscript is based and to answer expeditiously any questions about the manuscript or its authors.