Retina is the only part of human body where blood vessels and nerves can be directly observed. Through deep
learning and artificial intelligence, eye-si™ screens retinal images to diagnose more than 20 diseases in the
entire fundus, including glaucoma and maculopathy
*The first fundus disease assisted diagnosis software approved for application in China
The average incidence of positive signs of fundus diseases in the whole population is 26.99%, and that in the group above 60 years old is 68.95%.
In 2020, there were 37.3 million patients with diabetic retinopathy, 42.2 million patients with hypertensive retinopathy, and more than 100 million patients with eye diseases including retinal vein obstruction and macular degeneration in China.
In June 2020, the National Health Commission released the White Paper on Eye Health in China, which pointed out that there were only 44,800 ophthalmologists in China, and only about 1,000 of them were fundus specialists.
About 70% of diabetics never received standard fundus examinations. The outpatient fundus screen rate is less than 10%.
Applicable to hospitals, community clinics, physical examination centers, optometry centers, optical stores, pharmacies, insurance companies, etc.
Can be applied across departments, including endocrinology, ophthalmology, cardiology, nephrology, etc.
Real-world research at the cutting edge of global advancements：sensitivity 92.00%，specificity 96.86%
Based on deep learning technology, "eye-si" AI fundus screening system instrumentalizes and standardizes the diagnostic experience of fundus specialists, so as to improve fundus reading efficiency, reduce the probability of misdiagnosis and missed diagnosis, improve the diagnosis and treatment level and accuracy, and better realize the early detection, early intervention and early treatment of fundus diseases.
Integrate the functions of image acquisition, transmission, storage management, diagnostic inquiry and report management.
Based on deep learning technology, it carries out high-quality training in 100,000 + fundus image database and forms models, instrumentalizes the rich diagnostic experience of fundus specialists and magnifies the coverage by ten million times.
Regional organization device status, software license management, support information screening and classification, provide access management of multiple reading manufacturers, meet the needs of group hospitals or multi-level institutions.
Compatible for various fundus cameras, and image acquisition only takes 1.5 to 3 minutes
Fundus images enter AI reading center, and the diagnosis results will be issued within 20 seconds
(Manual review is added according to the scene) The final diagnosis report is issued and sent to the user end
Official account
