Axial and choroidal changes as key predictors of myopia control in red-light therapy: Evidence from machine learning models
Hu X, Wang Z, Li R, Sun B, Wang D, Zong H, Zhang X, Wei R, Jie H
Photodiagnosis and photodynamic therapy · 1 citation
Review labels
Neutral facts our review recorded about how this study was done. They describe method, never whether we like the result.
How it was studied
- Design
- Cohort study (indexed by PubMed)
- Studied in
- People
- Main outcome
- Health markers and function
Who paid for it
- Funding
- Funding not disclosed
Publication
- Published
- 2026-01-08 · Photodiagnosis Photodyn Ther · vol. 57 · p. 105342
- Publisher
- Elsevier BV
- Cited
- 3 citations · more than 99% of similar papers · 16.4× the field average
- Impact
- Top 10% most cited in its field
- References
- 46 works
- Access
- Open access (journal) · CC-BY-NC-ND
- Research areas
- Ophthalmology and Visual Impairment Studies · Retinal Diseases and Treatments · Corneal surgery and disorders
- Keywords
- Key (lock), Control (management), Foundation (evidence), Deep learning
- MeSH
- choroid, humans, myopia, sensitivity and specificity, cohort studies, adolescent, child, female, male, axial length, eye, machine learning, red light, boosting machine learning algorithms, predictive learning models
9 authors
From CN
- Xiaoxue HuWuhan Children's Hospital; Huazhong University of Science and Technology
- Zixun WangTianjin Medical University Eye Hospital; Tianjin Medical University
- Rui LiThe Central Hospital of Xiao gan
- Boxuan SunTianjin Medical University Eye Hospital; Tianjin Medical University
- D. X. WangWuhan Children's Hospital; Huazhong University of Science and Technology
- Hui ZongWuhan Children's Hospital; Huazhong University of Science and Technology
Abstract
Background
Repeated low-level red-light (RLRL) therapy has emerged as a promising non-invasive intervention for myopia control. However, the predictive factors underlying its efficacy remain insufficiently explored.
Methods
This multicenter cohort study included 538 pediatric patients who underwent RLRL treatment with a minimum follow-up of one year. Baseline ocular parameters and dynamic changes in axial length (AL) and choroidal thickness (ChT) over three months were collected. Multiple feature selection approaches-LASSO regression, Boruta, recursive elimination, and multivariate regression-were applied. Seven machine learning algorithms were trained, and their performance was evaluated using the area under the curve (AUC), sensitivity, specificity, and F1 score. SHAP and LIME analyses were utilized for interpretability.
Results
Logistic regression and gradient boosting models demonstrated the highest discriminative ability. XGBoost achieved optimal performance (AUC: 0.90-0.92; accuracy: 87.7-88.2 %; F1 score: 0.72-0.86). Across one- and two-year prediction models, six stable predictors were identified: age, baseline AL, anterior chamber depth (ACD), RNFL thickness at the temporal inferior quadrant (TI), ChangeAL, and ChangeChT. SHAP analysis revealed that ChangeAL was the dominant short-term predictor, whereas ChangeChT was most influential for long-term outcomes. Older age and greater choroidal thickening were consistently associated with a protective effect against myopia progression.
Conclusions
We developed and validated an interpretable machine learning model that accurately predicts short- and long-term outcomes of RLRL therapy in children. ChangeAL and ChangeChT serve as key dynamic biomarkers for treatment monitoring. These findings provide a foundation for personalized clinical decision-making in myopia management.
Abstract via Europe PMC. Copyright remains with the authors or publisher (CC BY-NC-ND).
Community trust
Loading…
How much do you trust this study's findings?
Comments
Sign in to rate, comment on or flag this study.Sign inSomething wrong here?
Flag this study if its information, labels or funding look wrong. An editor reviews every flag.
Sign in to rate, comment on or flag this study.Sign inEducational information about published research. Not medical advice, and not a recommendation to start or stop anything.