OPHTHALMIC IMAGE PROCESSING METHOD, OPHTHALMIC IMAGE PROCESSING DEVICE, AND OPHTHALMIC IMAGE PROCESSING PROGRAM

摘要

According to the present invention, an ophthalmic image of a subject eye to be evaluated is acquired, a plurality of subdivided images are extracted from the ophthalmic image, and the condition of the subject eye is predicted for each subdivided image on the basis of a trained model that has been trained in advance in relation to prediction of the condition of a subject eye, for each subdivided image, by means of machine learning performed by extracting a plurality of subdivided images from ophthalmic images for learning and employing correct answer data relating to the condition of each subdivided image, wherein the subdivided images are extracted from the ophthalmic image in such a way as to have a size corresponding to the condition of the subject eye to be evaluated.