The research team of the Health Science Center of the University of Texas in the United States has developed a device called "smart phone AI retinal tracker", which can complete retinal image analysis in one second, with an accuracy rate of 99%, bringing a breakthrough in the early screening of diabetes retinopathy. The team innovatively integrated AI applications into portable devices, using deep learning algorithms to enable professional level retinal image analysis even on ordinary smartphones. This technology not only enables ophthalmologists to improve screening efficiency, but also helps primary healthcare workers complete eye health examinations in routine diagnosis and treatment, making it particularly suitable for conducting high-quality retinal assessments in areas with scarce medical resources. The number of diabetes patients worldwide has exceeded 100 million, and diabetes retinopathy has become the primary cause of preventable blindness. To tackle this challenge, the team collected tens of thousands of retinal images covering different races and constructed a widely representative database. After verification, the outstanding advantages of this system are as follows: complete the staging diagnosis of diabetes retinopathy within one second, accurately distinguish diabetes retinopathy from other similar eye diseases, and its operation is simple and the cost is controllable. This technology can enable mobile medical devices to carry professional ophthalmic examination functions, and is expected to screen billions of people worldwide, fundamentally changing the pattern of prevention and treatment of diabetes eye disease. (New Society)