Intel claims that this AI-based solution can detect a variety of retinal conditions at the level of accuracy of human physicians. The company claims to be able to accurately identify diabetic retinopathy (DR), a condition in which the blood vessels in the back of the retina are damaged due to diabetes.
How the system works is really easy. Netra.AI analyzes images from portable, professional devices that use camera money, to get instant results for DR placement via a web-based portal. The solution uses AI algorithms developed in collaboration with advanced retina specialists with four deep steps of convolutional neural network (DCNN). This neural network helps in detecting the DR phase and defining lesions based on pixel density in fundus images.
One of the interesting facts about Netra.AI is that it is able to produce a complete report within two minutes of uploading images, enabling photographers to provide immediate advice to patients who need to be referred to a hospital. Intel claims the solution and delivers the best sensitivity and accuracy of 99.7% and 98.5% respectively.
Intel says the solution can be extended to other retinal conditions including glaucoma, a condition in which the optic nerve is damaged due to high eye pressure. The company says this will help reduce the burden of examining health care professionals and focus on essential services for patients in need of immediate care and intervention.
“Through Netra.AI, Sankara Eye Foundation and Leben Care have utilized the power of the Intel Xeon Scalable processor and the built-in Intel Deep Learning (DL) Boost to detect DR accurately and enable timely treatment to combat effective preventable vision and blindness in people with the disease. diabetic patients” said Prakash Mallya, vice president and executive director of Sales, Marketing and Communications Group, Intel India at the event.