18 August 2025
New research published in BioMedical Engineering OnLine has described a deep learning model for myopia severity classification from fundus images.
Researchers from the College of Biomedical Engineering and Anhui Institute of Optics and Fine Mechanics in China highlighted that the model classified images with an accuracy of 91%.
The scientists highlighted that classifying myopia severity using deep learning supports early screening and timely intervention.
“In resource-limited regions, such models enable large-scale screening and contribute to building a comprehensive myopia prevention and control framework,” they noted.
The model developed by the researchers is efficient, lightweight and has a reduced computational load – meaning that it suitable for deployment on a broad range of devices.
The scientists observed that the technology helps clinicians in adapting treatment to the needs of the individual.
“Since patients with different degrees of myopia require tailored treatment strategies, the model provides objective assessments to support individualised treatment planning and enables longitudinal analysis of treatment outcomes, thereby reducing unnecessary healthcare resource utilisation,” they explained.
Comments (0)
You must be logged in to join the discussion. Log in