Novosibirsk schoolboy Sergey Matveev, studying in the 11th grade at SUNC NSU, together with his supervisor, Yana Dementieva, a laboratory assistant at the Interdepartmental Laboratory of Engineering Design at SUNC NSU, have created an AI that can recognize retinopathy in infants. This is a lesion of the retina of the eyeball, resulting in optic nerve atrophy and irreversible blindness.
Now the disease is detected with the help of a lamp and magnifying glass, as well as laser scanning of the retina.
The method created by a Russian schoolboy allows to detect the disease from the images of the eye fundus. The young researchers created five different models for diagnosing blindness, each of which was tested on 20 thousand images of the eye fundus with different degrees of retinopathy.
As a result, the AI learned to recognize retinopathy with 98% accuracy.
In the future, the schoolboy plans to improve the model and create a graphical interface for medical professionals.
Scientists have proposed a new way of vaccination against diseases.