Korea Research Institute of Standards and Science (KRISS) said on the 29th that it developed an "artificial eye" that precisely replicates the structure of a real human eye. The technology is expected to significantly improve early diagnosis and treatment accuracy for retinal diseases by providing a standard to objectively evaluate and calibrate the performance of retinal diagnostic devices used in ophthalmology.
The retina is a vital tissue that detects light and transmits visual information to the brain. However, the number of patients with retinal diseases has steadily increased due to aging, increased use of electronic devices, and genetic factors. Retinal damage is difficult to recover, making early diagnosis and continuous monitoring especially important.
Currently, ophthalmology uses various devices, such as optical coherence tomography (OCT) and fluorescein angiography, to examine the retina, but accuracy has been limited because measurements differ by hospital and device, and there has been no standard to compare or calibrate them.
To address this, the KRISS Nanobio Measurement Group and the Medical Convergence Measurement Group jointly developed a "retina-mimicking eye phantom" that finely simulates the structure and function of a real human retina. The phantom reproduces 13 retinal layers, curvature, a microvascular network, blood flow, and retinal autofluorescence almost identically to a real eye. Its structural characteristics match more than 90% with an actual retina, and it can be applied to various diagnostic devices from tomography to angiography. The research findings were published in the July issue of Communications Engineering.
Using the phantom, core performance such as image resolution and field of view can be objectively verified, and differences in diagnostic results between devices can be reduced. This will allow both medical institutions and patients to obtain consistent and reliable retinal test results anywhere.
It is also expected to be used in industry and education. Device manufacturers can use the phantom to check prototype performance and strengthen quality control, and medical institutions can conduct training on device operation and diagnosis in environments similar to real patients.
Lee Sang-won, head of the KRISS Nanobio Measurement Group, said, "As AI technology expands in retinal disease diagnosis, this phantom will provide high-quality training data and also help improve the performance of AI-based diagnostic devices."
References
Communications Engineering (2025), DOI: https://doi.org/10.1038/s44172-025-00475-6