Caltech researchers develop algorithm to predict IQ scores from brain fMRIs
A new algorithm developed by researchers from the California Institute of Technology, Cedars-Sinai Medical Center in Los Angeles, and the University of Salerno in Fisciano, Italy, may be able to predict a person's intelligence quotient from fMRI brain scans.
Study participants were scanned with functional MRI (fMRI) while at a resting state. The algorithm then predicted IQ scores based on the images, said Ralph Adolphs, PhD, director and leadership chair of the Caltech Brain Imaging Center, in a prepared statement. The researchers chose to measure intelligence because its stability over a long period of time.
Aldolphs and colleagues trained the algorithm using data collected by the Human Connectome Project (HCP), an initiative funded by the National Institutes of Health (NIH) aimed at understanding the many connections in the human brain," according to the news release. Roughly 900 brain scans and intelligence scores from individuals were used to train the algorithm. The combined system proved successful in predicted intelligence levels for participants.
However, the scans are "coarse and noisy measures" of brain activity, with a lot of discarded information, explained Julien Dubois, PhD, a postdoctoral fellow at Cedars-Sinai Medical Center.
"The information that we derive from the brain measurements can be used to account for about 20 percent of the variance in intelligence we observed in our subjects," Dubois said, in a prepared statement. "We are doing very well, but we are still quite far from being able to match the results of hour-long intelligence tests, like the Wechsler Adult Intelligence Scale."
The researchers also tested the algorithm to ensure its accuracy by extracting more precise estimates of intelligence scores from 10 different cognitive tasks the subjects had taken, according to the news release. Ultimately, the researchers hope the availability of large data sets may one day make fMRI a promising diagnostic tool for conditions such as autism, schizophrenia and anxiety.
The research is currently waiting to be approved for publication by the Personality Neuroscience and Philosophical Transactions of the Royal Society, according to the news release.
A recent graduate from Dominican University (IL) with a bachelor’s in journalism, Melissa joined TriMed’s Chicago team in 2017 covering all aspects of health imaging. She’s a fan of singing and playing guitar, elephants, a good cup of tea, and her golden retriever Cooper.