DeepMind, an artificial intelligence unit, will use the data of up to a million eye scans to train its computers to identify eye defects, writes Bloomberg. The ultimate goal is to give doctors a digital tool that can read an eye-scan test and recognize problems faster. According to a statement by DeepMind, earlier detection of eye disorders related to diabetes and age-related macular degeneration can help doctors to prevent loss of vision in many people. The pilot project is conducted in collaboration with Moorfield’s Eye Hospital NHS Foundation Trust.
Google acquired DeepMind in 2014 to expand its artificial intelligence capabilities. It specializes in machine learning, the increasingly important area of technology where algorithms allow computers to learn and figure things out on their own.
Similar AI-project IBM Watson Health
Recently IBM announced a similar pilot project with its AI-platform Watson. IBM Watson Health believes it can help doctors to make a better diagnosis, by removing some of the human error. For example when radiologists looks too quickly at a mammogram and miss something, by using a machine that doesn’t get tired or confused. Reason enough for Watson Health to partner with over 15 hospitals and companies that are using imaging technology such as X-rays to capture information about body and health, to see how "cognitive imaging" works in the real world.
Health care review board
Separately, DeepMind has announced the creation of a health care review board that will scrutinize its work with the NHS. The nine-member panel includes a patient safety advocate, writes the editor of The Lancet medical journal, and others members with health and technology backgrounds. The reason for this step is that DeepMind’s relationship with the NHS has been criticized by some privacy advocates. They worry shared data will be used for other purposes besides medical advances. DeepMind assures will only be for health care purposes.