Ophthalmic diagnostic assistance system based on deep learning algorithm

Ophthalmic diseases have globally afflicted visual functioning and activities of daily living for many people. Among these vision-threatening disorders, glaucoma, diabetic retinopathy, and cataracts constitute major causes of visual deficit. Although preventative diagnosis could have precluded the development of vision loss in many individuals, suitable screening methods for early detection of ophthalmic diseases have not yet been explored. This study presents an ophthalmic disease diagnostic assistance system powered by deep learning algorithms, which can identify common ophthalmic conditions by interfacing with the Wechat Applet. The performance metrics of the proposed assistance system are a response speed of 20ms and an accuracy of 88.7% in diagnosing three common ocular diseases.