Apply Machine Learning Methods to Predict Failure of Glaucoma Drainage

The purpose of this retrospective study is to measure machine learning models' ability predict glaucoma drainage device failure based on demographic information and preoperative measurements. medical records 165 patients were used. Potential predictors included the patients' race, age, sex, intraocular pressure (IOP), visual acuity, number IOP-lowering medications, type previous ophthalmic surgeries. Failure was defined as final IOP greater than 18 mm Hg, reduction in less 20% from baseline, or need for reoperation unrelated normal implant maintenance. Five classifiers compared: logistic regression, artificial neural network, random forest, decision tree, support vector machine. Recursive feature elimination used shrink grid search choose hyperparameters. To prevent leakage, nested cross-validation throughout. With a small amount data, best classfier but with more classifier forest.