Exploring the Diagnostic and Prognostic Potential of OCT Data in Multiple Sclerosis

Exploring the Diagnostic and Prognostic Potential of OCT Data in Multiple Sclerosis

Machine learning (ML) is transforming Multiple Sclerosis (MS) diagnosis by leveraging Optical Coherence Tomography (OCT) data to assess retinal changes linked to neuro-degeneration. This study analyzed OCT data from 230 MS patients using ML models, including Random Forest (RF), Support Vector Machine (SVM), XGBoost, and k-Nearest Neighbors (KNN), to classify MS severity. RF emerged as the best performer, achieving F1-Scores of 0.74 (left eye) and 0.72 (right eye), with key retinal features such as the Superior and Temporal sectors, Central ILM-RPE, and asymmetry metrics identified as critical predictors. However, challenges remain, including measurement variability across OCT devices and segmentation inconsistencies, showing the need for standardization. Explore the modeling resources here