The diabetes diagnostic speed is very important. Delays in this case are fraught with complete blindness. The algorithm, which was developed by the Google and Verily specialists developed in one of the Indian clinics, is called upon to facilitate the work.
Diabetic retinopathy is a severe complication of diabetes mellitus, which affects the vessels of the eyeball. Manifests itself in every ninth sick and can really cause blindness.
Employees of the ophthalmological clinic of Arabind, located in the city of Madurai, worked on the technology of automatic detection of this pathology for 16 years. But only two years ago, in cooperation with Google engineers, they managed to achieve acceptable speed and accuracy.
The algorithm was taught the diagnosis of diabetic retinopathy in 128 thousand ophthalmological images of eyeballs. The effectiveness of the method was tested at 2 sets of 12 thousand pictures in each. The accuracy of the diagnosis has reached 97.5 percent. And this is faster and more accurately than the doctors.
A similar study was conducted by Google last year together with the Rajavithi State -Ench in Thailand. The results of the algorithm were then similar: the accuracy was 95 percent. And ophthalmologists were mistaken in a quarter of cases.
Google’s algorithm recently received a compliance with the standards of the EU EU CE. Thanks to him, doctors of the clinic managed to diagnose a larger number of patients with the appointment of the necessary treatment.
Source: Venturebeat