Brain study points way to possible imaging improvements

May 25, 2016
functional magnetic resonance imaging or fMRI brain
Changes in blood flow may not be as tightly tied to neuronal activity as previously thought.


When Prakash Kara, Ph.D., talks about his work on brain imaging, he uses an image that evokes the voyage of the Titanic. “Facing an iceberg, all you see is what’s above the water. But then there’s all this other stuff below it.”

The “other stuff” can make all the difference, whether it’s part of an iceberg big enough to bring down an “unsinkable” ship or the repertoire of neuronal responses in the brain, Kara’s specialty.

Kara, a neuroscientist at the Medical University of South Carolina, takes a deep dive into brain imaging in his latest study appearing this week in the prestigious scientific journal “Nature.” In a way, his approach to this complex project was simple. “Basically, I want to tackle a problem in a way nobody else has thought of,” Kara said.

The problem he tackles in “Nature” is, in part, to figure out what drives the signals recorded during hemodynamic imaging, including functional magnetic resonance imaging or fMRI, which is used widely in human neuroimaging studies. It’s thought to show where in the brain certain functions, such as seeing, hearing and speech happen. It does that by tracking changes in blood flow, causing areas to light up on scans when blood flow increases in those places. 

But Kara and his colleagues wondered if increases in blood flow are really tightly coupled to increases in neuronal activity. "Does the blood rush to the very specific part of the brain that is being used at that time or does it is spread out over a large area?” Kara asked. “Does the blood flow also increase in other areas that aren’t active?"

Knowing the answers could allow researchers and physicians to put fMRI images in better perspective and might help them diagnose and treat some medical conditions, including stroke and vascular dementia, more accurately.

“One long-term goal of the laboratory is to see whether it is worthwhile to do very high-resolution fMRI. Why spend all the money to do super high-resolution fMRI if it won’t get what you are after? A lot of people are saying that we just need a better magnet, and then we will get more detail. And our data are saying, ‘not necessarily,'” Kara said.

“When you present a stimulus, clusters of neurons respond and release something and that rapidly dilates a blood vessel locally. The dilation in turn increases blood flow. We wondered whether the neural and vascular signals matched in space and time.”

In this new study, they were expecting that the neurons and blood vessels would both respond to the same sensory stimuli, but his team noticed that they did not. “So why is it not matched? Why does the blood vessel dilate to other stimuli?” Kara asked.

Using animal models, Kara and his colleagues examined the mismatch more carefully. They found that while blood flow in individual vessels did increase in response to local neural activity, it also increased in response to sensory stimuli that did not evoke local neural activity near the blood vessel. Their publication shows that something else was at work.

"That meant that fMRI, while excellent at showing the general function of an area of the brain, may not be as good at direct neural recordings", Kara said. “The individual blood vessel response is much more promiscuous – a lot less selective – than the neural response. We didn’t know that before embarking on this study.”

That finding goes against conventional wisdom in his field. “The dogma was that we should see a change in blood flow, at least in the capillaries, that is very specific to what the local neurons are doing. Hence, there has been a tremendous drive that very high-resolution fMRI could give us neural circuit resolution.”


About the Author

Helen Adams