AI algorithm automatically diagnoses retinopathy of prematurity

A US research team published a paper in the latest issue of JAMA Ophthalmology, saying they developed an artificial intelligence (AI) algorithm that automatically diagnoses retinopathy of prematurity and has a diagnostic accuracy that is higher than most professional doctors.

Retinopathy of prematurity is a serious blinding eye disease caused by abnormal proliferation of retinal blood vessels and is common in premature infants. This disease is mainly diagnosed by doctors looking at the baby's eyes, subjective. In the United States, misdiagnosis of this disease is often caused by the lack of experienced clinicians.

To improve this situation, a team of researchers from the Oregon Health and Science University, Massachusetts General Hospital, the University of Illinois at Chicago, and the Columbia University Medical Center developed a new AI algorithm. The algorithm uses deep learning techniques to train through images of more than 5,000 baby eyes to learn how to distinguish between healthy blood vessels and diseased blood vessels. Test data showed that the accuracy of the algorithm for the diagnosis of retinopathy of prematurity was as high as 91%, and a team of eight premature retinal specialists diagnosed the same image with an accuracy of 82%. That is to say, in the diagnosis of retinopathy of prematurity, the AI ​​algorithm performs better than most professional doctors.

The test population in this study was mainly white infants. The research team is currently working with the Indian side to test whether the algorithm can accurately diagnose retinopathy of prematurity in India. In addition, the research team is still exploring whether the algorithm can detect abnormalities in other parts of the retina than in the blood vessels. The researchers said they hope the AI ​​algorithm will be used in clinical practice as soon as possible.

AI applications are currently becoming a hot research topic in the medical field. Last month, the US Food and Drug Administration (FDA) approved an AI device for detecting diabetes-related eye diseases, becoming a landmark event in the US for the application of AI technology in the medical field. (Technology Daily Liu Haiying)


Structural Bolts

YokeLink supply a vast range of Industrial Fasteners to DIN, ISO & ANSI standards, including: Nuts, Bolts, Screws, Washers .There are hundreds of combinations of materials, heads, threads, plating, heat treating, and secondary operations. The materials and finishes including Steel, Brass, Stainless Steel, Aluminum, Zinc plating, Galvanized and Black. Providing standard and specialised Fastener solutions, coupled with industry customers. We also have the flexibility, capacity and expertise to manufacture fully customised fasteners specific to your application.

bolts


Industrial Fastener, Hex Bolts, Screws, Flange Bolts, Structural Bolts

Ningbo Yokelink Machinery Co.,Limited , https://www.yokelink.com