Researchers from Australia, China and the United States have developed an artificial intelligence (AI) system that can detect and monitor retinal diseases on a larger scale.
Researchers from Monash University, Sun Yat-sen University, Beijing Eaglevision Technology, Beijing Tongren Eye Centre, Capital Medical University and the University of Miami Miller School have developed a new comprehensive AI system (CARE) for retinal diseases.
See also: Teenagers created the Acti-Vest vest for people with visual impairments
The CARE system, according to the researchers, was developed using fundus photography combined with a deep-learning system, which was trained using data from real-life case studies of retinal diseases.
Fundus photography is a procedure for taking photographs of the inside of the eye through the pupil, to examine retinal diseases.
“The CARE AI system has been trained to detect the 14 most common retinal using 207,228 fundus photographs from 16 clinical sites in Asia, Africa, North America and Europe,” said Associate Professor Zongyuan Ge.
See also: Backpack AI notifies visually impaired pedestrians of obstacles!
CARE has been tested using thousands of photos collected from 35 medical settings across China, including hospitals and physical examination centers.
CARE's performance was compared to that of 16 ophthalmologists and tested using datasets with non-Chinese ethnicities. From these tests, Ge said that the CARE system's performance was similar to that of professional ophthalmologists.
"These findings show that the AI system is accurate compared to the results of a professional and could allow more testing to be conducted on a larger scale," he said.
See also: Surgeries will become safer with the help of AI
The researchers hope to make CARE commercially available in China and later in the Asia-Pacific region , and plan to create a database of screening images from real-world environments that can be made available to hospitals to better diagnose retinal diseases.
“I hope that through this work we will be able to continue to see technological advances in this space,” said University of Wisconsin-Madison Imaging Diagnostic Center director Amitha Domalpally.
Source: ZDNet
