Researchers from the Hospital Universitario de Canarias (HUC) and the Universidad de La Laguna (ULL) have designed an artificial intelligence (AI) system capable of analyzing images of the fundus of the eye and detecting signs of glaucoma with a precision superior to 91%.
According to information provided by the Consejería de Sanidad del Gobierno de Canarias this Monday, the project has been distinguished with the Award for Best Work presented at the twentieth Congreso de la Sociedad Española de Glaucoma, recently held in Toledo.
"Its main contribution is the improvement of information during the decision process, allowing to visually identify the areas of the image that have been decisive in the diagnosis", adds the note, which details that the study is titled ‘Superpixel-based transformer vision: a new method to improve the explainability of Artificial Intelligence'.
Unlike other Artificial Intelligence models, usually considered as “black boxes”, this tool offers an interpretation "clear and understandable for specialists, respecting the morphology of anatomical structures", Sanidad has explained, which has added that this contributes to reinforcing confidence in its use and facilitates its incorporation into clinical practice.
Glaucoma constitutes the main cause of irreversible blindness worldwide, which is why its early detection is essential, the regional department has highlighted.
Thus, the application of technologies based on Artificial Intelligence is positioned "as a key resource to improve early diagnosis and patient care".