Related Experiment Video
Updated: Nov 11, 2025
07:20
Inducement and Evaluation of a Murine Model of Experimental Myopia
Published on: January 22, 2019
10.2K
Machine learning prediction of pathologic myopia using tomographic elevation of the posterior sclera
Yong Chan Kim1, Dong Jin Chang2,3, So Jin Park2
1Department of Ophthalmology, Incheon St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea.
Scientific Reports
|March 27, 2021
Summary
This study introduces a new quantitative index using key eye markers to detect pathologic myopia. This machine learning model accurately identifies patients with this condition, aiding in early treatment prioritization.
Area of Science:
- Ophthalmology
- Medical Imaging
- Machine Learning
Background:
- Qualitative fundus photography aids advanced pathologic myopia detection but struggles with early-stage classification due to subjective bias.
- Accurate classification of early-stage pathologic myopia is crucial for timely intervention and management.
Purpose of the Study:
- To develop and validate a quantitative index for detecting pathologic myopia using key ocular markers.
- To compare the diagnostic performance of the novel index against existing quantitative measures.
Main Methods:
- Utilized fovea, optic disc, and deepest point of the eye (DPE) to quantify relative tomographic elevation of the posterior sclera (TEPS) in 860 myopic patients.
- Employed a support vector machine (SVM) based machine learning classifier to predict pathologic myopia using TEPS, axial length (AxL), and subfoveal choroidal thickness (SCT).
Main Results:
- The TEPS index alone achieved an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.828 (77.5% sensitivity, 88.07% specificity).
- Combined TEPS, AxL, and SCT in the SVM model yielded excellent discriminative ability with an AUROC of 0.868 (80.0% sensitivity, 93.58% specificity).
- The combined model outperformed existing quantitative indicators (AUROC 0.758).
Conclusions:
- The developed quantitative index and SVM model offer an accurate method for identifying patients with pathologic myopia.
- This approach may assist in prioritizing patients for further treatment and management.
