To Speed Autism Diagnosis, Researchers Turn To AI
A public service announcement from Autism Speaks offers information about the signs of autism. (Ad Council for Autism Speaks/TNS)
A machine learning model can identify most children with autism before the age of 2, research suggests, potentially accelerating the process of getting a diagnosis.
In a study looking at data on about 12,000 people in the U.S. with and without autism, researchers said their artificial intelligence model dubbed AutMedAI could predict autism in children under age 2 with nearly 80% accuracy.
The model relied on 28 different parameters, all of which could be obtained without extensive assessments or tests, according to findings published this month in the journal JAMA Network Open.
Advertisement - Continue Reading Below
Some of the factors considered include the age when a child first smiled, when they said a short sentence and the presence of eating difficulties.
“The results of the study are significant because they show that it is possible to identify individuals who are likely to have autism from relatively limited and readily available information,” said Shyam Rajagopalan who led the study while at the Karolinska Institutet in Sweden.
While autism can be reliably diagnosed by age 2, many kids are much older before that happens, according to the Centers for Disease Control and Prevention. Early diagnosis is considered critical because it allows children to access intervention.
“This can drastically change the conditions for early diagnosis and interventions, and ultimately improve the quality of life for many individuals and their families,” Rajagopalan said of the study’s findings.
The researchers are now planning additional improvements — including adding genetic information to the model — and they want to study the approach in clinical settings.
“To ensure that the model is reliable enough to be implemented in clinical contexts, rigorous work and careful validation are required,” said Kristiina Tammimies, an author of the study from the Karolinska Institutet. “I want to emphasize that our goal is for the model to become a valuable tool for health care, and it is not intended to replace a clinical assessment of autism.”
Read more stories like this one. Sign up for Disability Scoop's free email newsletter to get the latest developmental disability news sent straight to your inbox.