Google successfully used machine learning to detect diabetic retinopathy, which is comparable to ophthalmologists

Diabetes itself does not necessarily cause harm, but long-term blood sugar can cause a variety of acute and chronic complications, which may lead to blindness, cardiovascular disease and renal failure, even lower limb amputation, and may even lead to death in severe cases. Among them, diabetic retinopathy (DR) is one of the most important manifestations of diabetic microangiopathy. The affected diabetic patients may have vitreous hemorrhage and may lose vision. Fortunately, Google has been trying to use machine learning to help diabetics detect diabetic retinopathy. And now this search giant has achieved success.

Google product manager and medical doctor Lily Peng said: "A few years ago, a Google research team began to explore the use of machine learning to screen for diabetic retinopathy (DR). Today, we are in the Journal of the American Medical Association (JAMA) The results of our research are published: A deep learning algorithm that interprets signs of DR in retinal photographs may help doctors screen more patients, especially in communities with limited resources."

Lily Peng explained: "The results show that the performance of our algorithm is comparable to that of an ophthalmologist. For example, on the validation set described in Figure B, the algorithm has an F-Score of 0.95 (a measure of comprehensive sensitivity and specificity, with a maximum of 1) The algorithm performance is slightly higher than the median F-Score (0.91) of the 8 ophthalmologists we consulted."

Taiwan switches

Induction Cooker Main Board,Motor Starter Button,Ignition Starter Switch For Foton,Main Spindle Induction Motor

Ningbo Zhenhai Rongda Electrical Appliance Co., Ltd. , https://www.centrifugalswitch.com