A Study on Visual Prognostic Estimation Method for Branch Retinal Vein Occlusion with Macular Edema

Abstract

In this study, we have investigated a method for estimating visual prognosis in branch retinal vein occlusion (BRVO) with macular edema (ME) using clinical data and machine learning. That is because ophthalmologists cannot predict individual patients’ visual functions. In addition, quantitative prognostic evaluation is required in order to obtain the informed consent of patients. In this paper, we carried out a binary classification, in which patients are classified into good or poor prognostic cases. 66 patients who were followed up for 12 months at Mie University Hospital were targeted. After we performed the feature extraction on clinical data, logistic regression analysis was employed to classify the patients’ prognosis. At the end of the experiment, we managed to achieve a mean of area under the curve (AUC) in the receiver operating characteristic (ROC) curve of approximately 0.92. In the future, we will design the classification model for practical use and attempt to optimize its parameters, including classification thresholds.