Deep learning-based automatic image quality assessment in ultra-widefield fundus photographs.

Summary: Doctors use special cameras to take wide pictures of the back of the eye. These pictures help find eye diseases. But if the picture is blurry or off-center, it is hard to use. Researchers built a smart computer program (AI) to check if these eye pictures are good enough. They tested the AI on thousands of pictures and compared it to three human eye doctors. The AI actually agreed with the doctors better than the doctors agreed with each other! This means the new computer program is a fast and reliable way to make sure eye pictures are high quality.