We are building a deep-learning powered image classification algorithm that is able to predict the likelihood of glaucoma given a human retina image. Our software utilizes the highly optimized Google image classifier named Inception v3, which has been trained to recognize the presence of multiple common items within an image. Using a training paradigm called transfer learning, we have re-trained the Inception classifier to instead recognize the shapes and patterns that are found in cases of retinal glaucoma. Our software will then be deployed to a website that can be accessed via an app on either an iPhone or Android mobile phone. We hope to show the feasibility of such a clinical decision support system, in which physicians are able to use the prediction of an algorithm that can be retrieved quickly and easily to inform their own decision on patient care.