A new breakthrough in the field of medicine has come to life. Researchers have just recently developed a system that predicts which babies will develop autism spectrum disorder (ASD) by the age of two, with an accuracy of 96 percent. The study, which was published in the journal Science Transitional Medicine, uses a combination of functional magnetic resonance imaging (MRI) and AI learning.
The study began by analyzing the brain scans of 59 younger siblings of children with ASD, as they are 20 times more likely to get autism than normal. Activity was captured and analyzed in around 230 regions of the brain of the six month old babies, which amounts to 26,335 neural connections. Then, once the children reached 2 years of age, they underwent behavioral analysis, and it was then that the children who actually developed autism were revealed (in this case, 11 of them did).
The scientists then used this data to feed on a machine-learning algorithm to look for unique patterns found in the then 6 month old autistic children. Once that was done, they had the system analyze the six month old brain scans again, and it was able to correctly identify 9 of the 11 children who had developed autism. (Know that the machine only remembers the patterns found in the brain scan, not the entire specific scans themselves).
This new method of identifying autism via AI opens up a whole new realm of diagnosis and treatment for ASD. Not only is it non-intrusive, it can also identify the warning signs of autism in just one scan, and has the potential to increase “the feasibility of developing early preventative interventions for ASD.”
“This is a game-changer for the field,” said Kevin Pelphrey, Director of the Autism and Neurodevelopmental Disorders Institute at George Washington University.
Source: New Scientist
This isn’t the first time AI systems have been developed for use in the medical field. It’s particularly useful in the diagnostic sector, since they can analyze data quickly and more accurately than a human. AI has been used to accurately depict which diseases 65 year olds are most likely to die from. IBM’s Watson has also already begun being used to diagnose diseases at an early stage, saving lives in the process.
It’s interesting to see the potential AI has in revolutionizing the healthcare industry. They are able to identify diseases earlier, faster, and more accurately than anyone ever could. This’ll help doctors perform cheaper and less intrusive preventive treatments on patients before they get any worse. It’s good to see technology is not just being built to make our everyday lives easier. It’s also being used to save lives.