Related Experiment Video
Updated: Jan 17, 2026
07:18
Evaluation of Capillary and Other Vessel Contribution to Macular Perfusion Density Measured with Optical Coherence Tomography Angiography
Published on: February 18, 2022
2.1K
Machine learning-based migraine analysis using retinal vessel diameters from optical coherence tomography: an
Fırat Orhanbulucu1,2, Metin Ünlü3, Duygu Gülmez Sevim3
1Department of Biomedical Engineering, Faculty of Engineering, Inonu University, Malatya, Türkiye.
Summary
Optical Coherence Tomography (OCT) retinal imaging can predict migraine. Machine learning models, particularly ensemble-based boosting, achieved high accuracy in distinguishing migraine patients from healthy controls.
Area of Science:
- Ophthalmology
- Neurology
- Medical Imaging
Background:
- Migraine is a complex neurological disorder affecting the central nervous system and retinal vasculature.
- Retinal imaging techniques, like Optical Coherence Tomography (OCT), offer a promising avenue for studying neuro-ophthalmological conditions.
- This study explores the potential of OCT-derived measurements for migraine prediction.
Purpose of the Study:
- To predict migraine using measurements from retinal images obtained via OCT.
- To evaluate the efficacy of machine learning algorithms in classifying migraine patients based on OCT data.
Main Methods:
- Examined 70 eyes of migraine patients and 38 eyes of healthy controls.
- Utilized features including retinal artery/vein diameters and choroidal thickness.
- Applied the SMOTE method for data balancing and Pearson's Correlation Coefficient for feature analysis.
Main Results:
- The LightGBM algorithm demonstrated superior performance in classifying migraine patients.
- Achieved high classification metrics: 93.28% AUC, 91.14% Accuracy, 86.67% F1-score, 0.74 Kappa statistic, and 0.76 Matthews Correlation Coefficient.
- Ensemble-based boosting models outperformed traditional machine learning classifiers.
Conclusions:
- Machine learning algorithms show effective performance in predicting migraine from OCT data.
- This preliminary study highlights the potential of OCT imaging in migraine diagnosis.
- Ensemble-based boosting models are particularly effective for this application.
