2023-12-19 15:17:16
Among the medical applications that artificial intelligence (AI) can provide, a multi-institutional Korean team of child and adolescent psychologists, behavioral scientists, ophthalmologists and biomedical systems informatics specialists has discovered that AI based on Deep learning can accurately diagnose children with autism.
In his study, published in ‘JAMA Network Open‘, the group has designed an autism diagnostic tool using a learning-based artificial intelligence system and tested it on people.
He autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by problems with communication, social interaction, and often repetitive movements.
Previous research has suggested that up to 1 in 36 people have some degree of autism. Additionally, it is known that the sooner it is diagnosed, the more can be done to help people with this disorder overcome the problems that many experience.
Based on the fact that some disorders like ADHD can cause minor retinal abnormalities, the researchers conducted an experiment to determine if the same thing also occurred in ASD.
The team taught an artificial intelligence system to recognize patterns in the retinas of children with autism. The system was then used to scan retinal images of 958 children, half of whom had the disorder, and to determine which of the children had autism.
The system found each of them and did not produce any false positives.
Clinical trial
Additionally, he assigned a numerical score that gave an estimate of where on the spectrum those identified with ASD fell. In this case, however, it was not as accurate, as the accuracy rate was 48% to 66%, compared to assessments done using standardized tests like the Autism Diagnostic Observation Schedule.
In their experiment, the children involved were between 4 and 18 years old; therefore, it is not yet known whether the system would be as accurate with younger children (the retina does not fully develop until age 4).
#test #detect #autism #early