Machine learning analysis with the comprehensive index of corneal tomographic and biomechanical parameters in detecting pediatric subclinical keratoconus

Aug 5, 2026

Early diagnosis of pediatric keratoconus is critical because clinical focus is often on adults, yet timely detection can prevent progression and improve patients’ quality of life. This case-control study applied machine learning to assess corneal tomographic and biomechanical variables for detecting subclinical keratoconus in pediatric eyes, comparing KC, SKC, and age- and gender-matched controls. Diagnostic performance was evaluated with ROC analyses and model-based logistic regression using stepwise variable selection to reduce overfitting.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Background

Methods

Study design and participants

Statistical and machine learning analyses

Results

Diagnostic performance of parameters

Machine learning models and cutoff values

Abbreviations

TYPE Original Research PUBLISHED 06 December 2023 DOI 10.3389/fbioe.2023.1273500 OPEN ACCESS EDITED BY Matthew A. Reilly, The Ohio State University, United States REVIEWED BY Yingxue Zhang, Wayne State University, United States Fulvio Ratto, National Research Council (CNR), Italy Junjie Wang, Wenzhou Medical University, China *CORRESPONDENCE Kaili Yang, [kelly1992abc@163.com](kelly1992abc@163.com) RECEIVED 06 August 2023 ACCEPTED 15 November 2023 PUBLISHED 06 December 2023 CITATION Ren S, Yang K, Xu L, Fan Q, Gu Y, Pang Cand Zhao D (2023), Machine learning analysis with the comprehensive index of corneal tomographic and biomechanical parameters in detecting pediatric subclinical keratoconus. Front. Bioeng. Biotechnol. 11:1273500 . doi: 10.3389/fbioe.2023.1273500 COPYRIGHT © 2023 Ren, Yang, Xu, Fan, Gu, Pang and Zhao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Machine learning analysis with the comprehensive index of corneal tomographic and biomechanical parameters in detecting pediatric subclinical keratoconus Shengwei Ren, Kaili Yang*, Liyan Xu, Qi Fan, Yuwei Gu, Chenjiu Pang and Dongqing Zhao Henan Provincial People ’s Hospital, Henan Eye Hospital, Henan Eye Institute, People’s Hospital of Zhengzhou University, Henan University People’s Hospital, Zhengzhou, China Background: Keratoconus (KC) occurs at puberty but diagnosis is focused on adults. The early diagnosis of pediatric KC can prevent its progression and improve the quality of life of patients. This study aimed to evaluate the ability of corneal tomographic and biomechanical variables through machine learning analysis to detect subclinical keratoconus (SKC) in a pediatric population. Methods: Fifty-two KC, 52 SKC, and 52 control pediatric eyes matched by age and gender were recruited in a case-control study. The corneal tomographic and biomechanical parameters were measured by professionals. A linear mixed-effects test was used to compare the differences among the three groupsandaleast signiﬁcant difference analysis was used to conduct pairwise comparisons. The receiver operating characteristic (ROC) curve and the Delong test were used to evaluate diagnostic ability. Variables were used in a multivariate logistic regression in the machine learning analysis, using a stepwise variable selection to decrease overﬁtting, and comprehensive indices for detecting pediatric SKC eyes were produced in each step. Results: PE, BAD-D, and TBI had the highest area under the curve (AUC) values in identifying pediatric KC eyes, and the corresponding cutoff values were 12 μm, 2.48, and 0 . 6, respectively. For discriminating SKC eyes, the highest AUC (95% CI) was found in SP A1 with a value of 0.84 (0.765, 0.915), and BAD-D was the best parameter among the corneal tomographic parameters with an AUC (95% CI) value of 0 . 817 (0 .729, 0 . 886) . Three models were generated in the machine learning analysis, and Model 3 (y = 0.400*PE + 1.982* DA ratio max [2 mm]−0.072 * SP A1−3.245) had the highest AUC (95% CI) value, with 90.4% sensitivity and 76.9% speciﬁcity, and the cutoff value providing the best Youden index was 0 .19. Abbreviations: KC, Keratoconus; SKC, Subclinical keratoconus; ROC, Receiver operating characteristic; BAD-D, Belin Ambrosio enhanced ectasia total deviation index; CDVA, Corrected distance visual acuity; K1 F, Flat keratometry; K2 F, Steep keratometry; Ka, Corneal keratometric astigmatism; Kmax F, The maximum keratometry; Kmean F, Mean keratometry; ACT, Corneal thickness at the pachy apex; PCT, Corneal thickness

FAQs

What problem does the study address?

The study targets the need for earlier detection of pediatric subclinical keratoconus, since keratoconus diagnosis has traditionally focused more on adults.

How were the pediatric groups compared in the analysis?

A case-control design compared pediatric eyes from KC, SKC, and age- and gender-matched controls, using measured corneal tomographic and biomechanical parameters.

Which indices and models showed the best diagnostic ability?

For identifying pediatric KC, PE, BAD-D, and TBI showed the highest AUCs. For discriminating SKC, SP A1 had the highest AUC, and Model 3 combining multiple parameters achieved the best overall performance with reported sensitivity, specificity, and a Youden-index cutoff.

Aug 5, 2026

Upload by Lucas Martin