Moving picture: A typical approach for analyzing functional connectivity differences between autistic and non-autistic children shows fewer group differences (left) than a new statistical approach based on machine learning (right).
Young autistic children and children with prominent features are often excluded from functional magnetic resonance imaging (fMRI) studies because they move their heads too much during scans, according to new research†
Applying a “missing data” approach that combines statistical modeling with machine learning may help address the problem, the study also finds.
Findings suggest results of autism imaging studies are biased, says Mary Nebelthe lead researcher and research scientist at the Kennedy Krieger Institute in Baltimore, Maryland.
Because researchers often discard fMRI data from participants whose movements cross a threshold, “we know we’re capturing some unrepresentative subgroups of children with an autism diagnosis,” says Kami Koldewynreader in psychology from Bangor University in Gwynedd, Wales, who was not involved in the work.
“It’s really nice to see such confirmation of something that we all know in the field to some degree,” she says of the new work.
NEbel and her colleagues analyzed resting-state fMRI scans of 545 children ages 8 to 13. None of the children has an intellectual disability and 148 of the 173 autistic children are boys. The researchers also measured the children’s characteristics of autism and attention deficit disorder/hyperactivity using the Autism Diagnostic Observation Schedule and standard parent questionnaires.
Two common quality control metrics for head movements had excluded more autistic than non-autistic children, the researchers found. The less selective metric excluded about 29 percent of autistic children, compared with 16 percent of non-autistic children, and the stricter measure excluded 81 and 60 percent, respectively.
Excluded children tended to be younger and had more prominent cognitive and social problems and poorer motor control. Each factor was associated with differences in functional connectivity, the researchers found, supporting their hypothesis that patterns in the types of participants excluded from fMRI datasets could influence the conclusions scientists draw about connectivity in the brains of people with autism. . The work was published in May in neuroimage.
CAusal inference, a way of analyzing data, can help correct biases in large observational studies in which “missing data is a big problem,” Nebel says.
By applying a ‘propensity’ causal inference model to the fMRI data, Nebel and her colleagues characterized the relationship between a participant’s traits and the usefulness of that person’s data; an ‘outcome’ model captured the relationship between a participant’s traits and their functional connectivity.
The two-pronged approach essentially served to increase the weight of useful data from underrepresented children in the sample, Nebel says. “Once you fit the model, you can extrapolate to make predictions about children who didn’t have useful functional connectivity data.”
For the fMRI data collected using the less selective motion criteria, the bias corrections were small, the authors wrote. The tightened criteria resulted in too few autistic children with usable data, which meant that they were unable to test their approach. The stricter criteria would likely have introduced greater bias and therefore led to greater bias corrections using their approach, Nebel and her colleagues wrote.
“It’s a very important paper,” says Max Bertoleroa former research scientist who is a neuroinformatics software architect at Nous imaging, a Missouri-based medical imaging software company. “Even if that’s not the final solution, at least they suggest something and see how well it works.” Bertolero was not involved in the investigation.
Still, “there’s quite a bit of evidence in the field that kids with autism are quite headstrong,” Koldewyn says. When it comes to brain changes, researchers must assume that the relationships between functional connectivity and other traits of high-motion children who pass quality control will be the same across all subgroups. “We don’t know,” she says.
AWhile approaches like those of Nebel and her team are new and important, they don’t replace efforts to reduce movement during fMRI studies in the first place, Bertolero says. “It can be part of a toolkit for dealing with movement, in addition to collecting more data, which is very expensive.”
Nous Imaging, for example, has had some success combining biofeedback and engaging games to help participants remain still in an fMRI machine. Koldewyn has also discovered that short breaks over the course of imaging can reduce how much a participant moves during the scans.
Post-imaging approaches and strategies to reduce movement during imaging are both important, Nebel says, not only for children with autism, but also for people with other conditions who tend to move during scans.
Cite this article: https://doi.org/10.53053/ZLQB7192
#Causal #Inference #Method #Reduces #Movement #Bias #Autism #Imaging #Research #Spectrum #Autism #Research #News