The AI test detects glaucoma at a very early stage – 18 months earlier than doctors do now. The development of scientists from University College London is described in the journal Expert Review of Molecular Diagnostics.
Glaucoma is the leading global cause of irreversible blindness, which affects more than 60 million people worldwide annually. Scientists predict that by 2040 this figure will double. This disease occurs as a result of cell death in the retina, the back of the eye.
Now, to detect the disease, a test called DARC is used, which involves the introduction of a fluorescent dye into the bloodstream, which binds to the cells of the retina and “highlights” the dead cells in bright white. The more spots of this color in the picture, the higher the risk of glaucoma.
The main problem with this method is that different doctors read pictures differently. To solve this problem of discrepancies, the researchers created AI, which allows us to uniquely determine the presence or absence of the disease.
The first two phases of the DARC AI clinical trials, in which 60 volunteers participated, determined the presence of the disease in 20 patients 18 months before it began. This is much faster and more accurate than now, the authors of the study note.