Traditional mydriatic fundus examination is time-consuming and labor-intensive. Many peripheral retinal lesions need early detection and diagnosis. The imaging range of 200°-240° is more than 5 times that of traditional fundus photography (45°), which can greatly improve the diagnosis rate of lesions. The team of Zhongshan Ophthalmic Center of Sun Yat-sen University based on more than 100,000 cases of ultra-widefield fundus colorpnods, developed an eye disease screening guidance system that based on artificial intelligence. The system can automatically and quickly screen many kinds of fundus lesions including retinal lattice degeneration, retinal hole, retinal detachment, retinal hemorrhage, retinal exudation, vitreous wart, etc. And can give corresponding medical guidance according to the characteristics of the lesion to prevent the progression of the lesion and save the visual function of the patients as much as possible. The result was published in Nature: Communications Biology January 2020.
The article link:https://www.nature.com/articles/s42003-019-0730-x
Figure 1: The screen-shot of the article