New brain map more than doubles its defined areas

Imaging technology can be important for more than diagnosing or treating individuals—a new study published in the journal Nature shows the abilities of MRIs to facilitate new discoveries about the human body.

The study’s results outline 97 new areas of the brain that scientists hadn’t been able to identify before. The information helps make more concrete identifications of the estimated 50 to 100 areas per brain hemisphere, 83 of which were already known.

The authors said they used “major advances” in MRI technology to map out all four neurobiological properties (architecture, function, connectivity, topography), plus the brain’s activity during certain tasks, throughout the entire brains of 210 people. Those images came from the National Institutes of Health (NIH) initiative called the Human Connectome Project. The maps were fed into programs that could pick out patterns and similarities among the individual scans, showing places where the brains had similar delineations.

To get these especially precise views of the brain, the analysts also needed to cut down on the signal-to-noise ratio of the MRIs without using spatial smoothing, which could have blurred some of the specificities they were looking for. That’s where area-wide scans of the brain came into play.

“An areal parcellation enables area-wise analyses (averaging data within each area), thereby improving SNR and statistical power without the deleterious effects of spatial smoothing (to the extent that properties within an area are uniform). Parcellation dramatically reduces data dimensionality,” the authors wrote.

The images showed previously unseen distinctions in nearly 100 newly defined areas.

Using a large sample size also allowed the researchers to identify which individual abnormalities they saw were part of normal human variation and which participants had actual atypical brain structures.

Ultimately, the computer program designed to read the scans identified 96.6 percent of the newly defined areas in the individual MRIs.

The new information could help physicians better make sense of individuals’ brain scans in the future. If they know roughly what to look for, it’ll be easier to see possible anomalies that could indicate need for certain diagnosis or treatment. Plus, researchers will be able to make population-wide assumptions about the brain more precisely using this map.

Tim Behrens, a contributor to the Human Connectome Project (but who did not work on this particular research, told the BBC, “It will now be the parcellation that is used by all of neuroscience, I would think."

But there’s still a lot left to learn.

The study’s lead author Matthew F. Glasser told the New York Times, “This map you should think of as version 1.0. There may be a version 2.0 as the data get better and more eyes look at the data. We hope the map can evolve as the science progresses.”