Artificial Intelligence (or AI) is increasingly being used in healthcare settings to improve patient diagnosis. While it was once thought of as a far-away future possibility, AI is now here, and it’s starting to make a pretty big impact…
Could artificial intelligence systems now become an integral part of the medical process, and help save the sight of millions of people?
The signs are looking positive.
What is AI?
While artificial intelligence often conjures up images of robots, the reality is a little more bland. This is an easy-to-understand definition:
“AI is using computers to solve tasks that would normally require the intelligence of a human”
The abilities that are being developed are vast but here are some key traits of AI:
- the ability to recognise voice commands.
- the capacity to study images and video.
- the capability to make informed decisions.
In the past few years, these human-like capacities have proved popular – AI systems have been successfully introduced into a number of industries.
One example are fatigue monitoring systems for vehicle drivers. Software is used to ‘read’ the driver’s pupils and assess their current state of tiredness. Large savings have been made through accident reduction and lower insurance premiums.
The medical field has been a little slower to react, yet the potential advantages of using artificial intelligence are massive. Imagine if a computer could diagnose disease as well as a human – waiting list times would be slashed and costs for huge organisations like the NHS would be dramatically reduced.
With these kinds of benefits on offer, it is no wonder that in the last few years, researchers have been concentrating heavily on developing and testing AI software.
The Google Game-Changer
A division of the Google Deep Mind team, comprising of computer experts and medical doctors, have been focusing their attention on diabetic retinopathy. 415 million people worldwide are at risk of developing this disease, so any screening improvements are extremely welcome.
Google researchers built an AI system and then tested it on over 12,000 images to see if it could accurately diagnose this condition. The results were remarkable – the computer had a higher success rate than individual doctors!
These findings are especially good news for rural communities and developing countries with fewer medical professionals. Finding a low-cost and more accessible screening solution could transform the early diagnosis of diabetic retinopathy in these areas.
What Other Research Has Been Done?
While Google’s project has been the most high profile, it is by no means an isolated case. The technology has had a number of initial successes, with studies showing that artificial intelligence systems can deliver a high level of competency.
Scientists recently used AI to see if it could be used in the detection of glaucoma. Like diabetic retinopathy, early detection is the key in fighting this disease.
Glaucoma researchers studied 106 eye images from 53 people who were at various stages of glaucoma disease. The outcome was again encouraging; 100% accuracy was achieved!
Elsewhere, ophthalmologists tested their wits against a machine when diagnosing cataracts. Both the medical professionals and the computer managed to attain the same level of success. Another great indicator for the future!
The Next Steps
As the technology is further refined, how will it actually be delivered? After all, it is one thing developing artificial intelligence software, but quite another installing it where it’s most needed.
The first possibility is to incorporate it into existing medical facilities. The systems could be used to assist doctors in diagnosing patients. Obviously, doctors could also oversee the accuracy of the software and troubleshoot any errors that may occur.
Another, and potentially more exciting prospect is using AI in smartphones.
Recent news suggests that most of the leading smartphone manufacturers are actively investing in artificial intelligence for their devices, which raises the intriguing possibility of self-diagnosis.
Will it really be possible to use a smartphone to take a photo of your eyes and assess their health? This would truly revolutionise eye treatments and could allow patients to skip the consultation phase.
In reality, it is hard to say exactly what will happen, but there is no doubt about one thing – artificial intelligence is about to cause a healthcare revolution, whether we like it or not.