Disease Diagnosing Google’s Deep Learning Technology Have Better Accuracy than Doctors

National Eye Institute on Flickr / This probably doesn't look like a diseased eye to you, unless you're an ophthalmologist or Google's new deep learning machine.

Boffins at Google have been busy working with its deep learning technology in an attempt to diagnose certain diseases. Also, because it’s continually learning, the system can now diagnose some diseases more accurately than most human doctors. One area that the robots are excelling in is in diagnosing diabetes. This is large because one of the problems that patients with diabetes often suffer with is a condition called diabetic retinopathy, and the machine learning robots have learned quite quickly from photos how to diagnose it accurately.


Google’s deep learning technology is capable of learning almost anything is you give it enough data and the right algorithm, and being able to detect diabetic retinopathy early on could save millions of people from becoming blind unnecessarily. One of the problems that many individuals have when it comes to this disease is that they don’t have a qualified ophthalmologist to go to who can diagnose the issue.

The team at Google worked with the deep learning technology to train it to recognize both healthy and diseased eyes. The results showed that loaded with the correct algorithm, the AI achieved a better accuracy level at diagnosing diseases than qualified ophthalmologists.  This type of technology is already being used to diagnose other diseases such as heart failure and certain types of cancer but isn’t quite as advanced as Google’s machine.


This is just another step forward for AI becoming integrated into society and as long as good results are being achieved, then why not welcome it in?  This is a breakthrough for diabetes patients. Hopefully, soon, they will not need to worry if the right professional is nearby as they can simply send their photo straight to the AI to get an accurate diagnosis.  Others that are also working on similar projects include IBM with its skin cancer detection algorithm and a Finnish group who are developing a malaria detecting algorithm.


More News To Read

Comments

comments