AMD, the country’s most prevalent cause of sight loss, manifests as “dry” in most patients. However, about 15 percent of patients experience “wet” AMD, which is more sight-threatening and occurs when abnormal blood vessels appear underneath the retina. This can lead to permanent vision loss if not treated early.
The AI system involves two deep convolutional neural networks, one inputting raw 3D scans and the other taking a segmentation map detailing the types of tissue in the retina. The prediction tool uses this data to estimate a patient’s risk of developing wet AMD.
The time frame before wet AMD develops is crucial for preventive care, and Google’s AI tool has shown potential to predict whether a patient with wet AMD in one eye will develop it in their second eye within six months.
Read the full news release here.
More articles on artificial intelligence:
Elon Musk claims AI chip could ‘fix anything that’s wrong with the brain’
Mount Sinai pioneers AI tech for rapid COVID-19 diagnosis
What’s on the AI wish list for 3 healthcare execs