Related Experiment Video
Updated: Jan 9, 2026
07:45
Quantifying Intermembrane Distances with Serial Image Dilations
Published on: September 28, 2018
6.7K
3D volumetric analysis outperforms 2D linear measurements in predicting vision after surgery for epiretinal
Myrta Lippera1,2, George Moussa1, Luke Michaels1,3
1Manchester Royal Eye Hospital, Oxford Road, Manchester, M13 9WL, UK.
Summary
Three-dimensional retinal measurements, particularly foveal volume, are superior to traditional two-dimensional parameters in predicting visual outcomes after epiretinal membrane surgery.
Area of Science:
- Ophthalmology
- Medical Imaging
- Machine Learning
Background:
- Epiretinal membranes (ERMs) can impair vision.
- Predicting post-surgical visual acuity is crucial for patient management.
- Traditional 2D measurements may not fully capture retinal changes.
Purpose of the Study:
- To analyze pre-operative biomarkers for predicting functional outcomes after ERM surgery.
- To compare the predictive value of 2D and 3D retinal measurements.
- To develop a machine learning model for visual outcome prediction.
Main Methods:
- Machine learning (XGBoost) was used to develop a predictive model.
- Pre-operative clinical and anatomical parameters, including 2D and 3D retinal measurements, were analyzed.
- Association between pre-operative parameters and post-operative best corrected visual acuity (BCVA) was assessed.
Main Results:
- Foveal retinal volume was the strongest predictor of post-operative BCVA.
- 3D retinal layer volumes showed greater predictive value than 2D measurements.
- Non-linear relationships were observed between key features and visual outcomes.
Conclusions:
- 3D retinal models offer greater utility than 2D parameters for predicting visual outcomes post-ERM surgery.
- Machine learning algorithms can effectively predict visual results.
- The study questions the influence of certain linear parameters on post-ERM surgery visual outcomes.
