Related Experiment Video
Updated: May 12, 2025
07:44
In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography
Published on: July 24, 2020
2.8K
Retinal and choroidal microvasculature and structural analysis in OCTA for refractive amblyopia diagnosis using
Xinlong Liu1, Caihong Xue1, Mengdi Li2
1Clinical College of Ophthalmology, Tianjin Medical University, Tianjin 300020, China; Tianjin Key Lab of Ophthalmology and Vision Science, Tianjin Eye Institute, Tianjin Eye Hospital, Tianjin 300020, China.
Journal of Optometry
|May 7, 2025
Summary
Adolescent amblyopia shows thicker retinal and choroidal layers. Machine learning on OCTA images accurately detects these changes, aiding in diagnosis.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Amblyopia, or 'lazy eye,' affects visual development in children.
- Current diagnostic methods for amblyopia may not capture subtle microcirculatory changes.
Purpose of the Study:
- To compare retinal and choroidal microcirculation and structure in amblyopic versus healthy adolescents.
- To develop a machine learning model for amblyopia classification using OCTA imaging.
Main Methods:
- Optical coherence tomographic angiography (OCTA) scans of the macula were performed on 19 adolescents with hyperopic refractive amblyopia and 22 controls.
- Retinal thickness, choroidal thickness, and capillary plexus perfusion densities were analyzed.
- Machine learning algorithms, including Random Forest and cross-validation, were applied for classification.
Main Results:
- Amblyopic eyes exhibited significantly increased retinal and choroidal thickness, especially in central and nasal regions.
- No significant differences in superficial and deep capillary plexus perfusion densities were found.
- The machine learning model achieved 92% accuracy in classifying amblyopic and normal eyes.
Conclusions:
- Refractive amblyopia is associated with thicker retinal and choroidal layers.
- OCTA combined with machine learning provides a robust framework for diagnosing refractive amblyopia.
- Automated classification using Random Forest and cross-validation enhances diagnostic precision.
