Summary: The different symptoms of autism spectrum disorder can affect different brain regions and neuroanatomy. Artificial intelligence technology has allowed researchers to discover brain differences at the individual level in people with autism.
source: Boston College
Differences in behavior between people with autism spectrum disorder (ASD) closely correlate with differences in neuroanatomy — the shape of the brain — a team of Boston College neuroscientists report today in the journal Sciences.
This discovery could help understand the causes of ASD, and develop personalized interventions.
The team used artificial intelligence (AI) to study MRI data from more than 1,000 individuals with autism and compared those images to AI-generated simulations of what the brains would look like if they did not have ASD.
“We found that different people with autism can be affected by different brain regions, and thanks to AI simulated brains, we were able to identify specific brain regions that differ between ASD individuals,” said Aidas Aglinskas, a postdoctoral researcher at Boston College. He co-authored the report.
“In addition, separating autism-related variance in brain anatomy from unrelated variance revealed subtle relationships between individual differences in brain anatomy and symptoms.”
Autism varies – in both symptoms and neuroanatomy – from one individual to another. Previous research hypothesized that there may not be a single set of neuroanatomical correlates common to all individuals with autism.
Confirming these suggestions has been difficult, Aglinskas said, because identifying autism’s neuronal alterations is such a challenging task. Brains differ due to many factors, including genetic variation not due to ASD, which is difficult to control in a research study.
Aglinskas, who conducted the research with Boston University associate professors of neuroscience Joshua, said the team crossed that barrier by using AI to identify patterns of autism-specific neural variance, which then allowed the team to identify neural pathways specifically affected in ASD. Hartshorne and Stefano Azzotti.
“Autism-related differences in brain anatomy can disappear between non-autistic differences,” said Aglinskas. As a result, it has been difficult to identify differences in brain anatomy associated with differences in symptoms. We have used artificial intelligence to separate autism-related differences from unrelated ones. “
The team set out to determine whether features specific to autism’s brain anatomy differ across individuals in a way related to their symptoms.
Previous studies examining individual differences in brain anatomy within ASD have not separated autism-specific traits from other unrelated individual differences in neuroanatomy, making it difficult to study the relationships between neuroanatomy and symptoms, Aglinskas said.
Using MRI data from 1,103 study participants, the team used an analytical method very similar to “deep fakes” — simulated images, videos, and other images created with patterns of visual data involving study participants are difficult to detect, according to the report. .
The team instead used computer-detected patterns to create a simulation of what each ASD individual’s brain would look like if they did not have ASD. The team reports that this is enabled by a new artificial intelligence technology, which separates individual differences in brain anatomy into autism-specific and unrelated features.
“We were surprised to find that although a great deal of variation in brain anatomy was observed between ASD individuals along multiple dimensions, the individuals did not cluster into distinct and categorical subtypes as previously thought,” Aglinskas said.
At the level of brain anatomy, individual differences within ASD may be better captured by continuous dimensions rather than categorical subtypes, but more importantly this does not rule out the possibility of finding categorical subtypes with other types of brain measurements, such as functional imaging. “
Going forward, the researchers point to the need for a more detailed understanding of how these neuroanatomical differences affect behaviour.
Anzlotti said the team plans to use AI tools to look beyond just brain structure for ways to better understand autism diagnoses and behavior of individuals with autism.
“Two brains can be formed very similarly but still function differently,” Anzlotti said.
There are a number of other aspects of the brain that we will need to look at to get a complete picture. For now, we’re focusing on functional connectivity – a measure of how the brain is “wired.” The big question is whether this will show us something new about individual differences within ASD. The goal of this type of work is to be able to use brain imaging data to help develop personalized healthcare approaches for those with autism spectrum disorder. “
About this research news for AI and ASD
author: press office
source: Boston College
Contact: Press Office – Boston College
picture: The image is in the public domain
original search: Access closed.
“Contrastive machine learning reveals the structure of neuroanatomical variance within autism” by Idas Aglinskas et al. Sciences
Contrastive machine learning reveals the structure of neuroanatomical variation within autism
Autism spectrum disorder (ASD) is highly heterogeneous. Identifying systematic individual differences in neuroanatomy can inform personalized diagnoses and interventions.
The challenge is that these differences are intertwined with difference for other reasons: individual differences not associated with autism and measurement purposes.
We used deep learning covariance to separate autism-specific neuroanatomical variance from covariance with typical control participants. Autism spectrum disorder-specific variation associated with individual differences in symptoms.
The structure of this autism-specific variance (ASD) also addresses a long-standing debate about the nature of autism spectrum disorder: at least in terms of neuroanatomy, individuals do not group into distinct subtypes; Instead, they are organized according to continuous dimensions that affect distinct groups of regions.