How does the communication between different brain regions produce complex cognitive abilities?

In the '90s, functional magnetic resonance imaging (fMRI) revolutionized the way we understood
the brain.
This non-invasive tool has become a backbone of neuroscience research, allowing scientists to delve deeper into neural connections in the brain and help us understand brain function and pathology
.

However, the current fMRI also faces its own bottlenecks
.
For example, this technology cannot answer a key question: How does communication between different brain regions produce complex cognitive abilities? The key factor contributing to this bottleneck is the lack
of temporal and spatial resolution of fMRI detection.

Now, a new study published in the journal Science promises a revolutionary breakthrough
.
Based on fMRI, the research team from South Korea has increased the temporal resolution of the detection to the millisecond level, so as to achieve direct imaging
of neural activity in the brain.

What's even more surprising is that the new technology, called direct imaging of neuronal activity (DIANA), doesn't use any entirely new instrumentation, but rather a major breakthrough
based on software-based improvements alone.

Professor Jang-Yeon Park of Sungkyunkwan University in South Korea was one of the leaders of
the study.
"I was shocked by advances in neuroscience, but at the same time, there was a question that bothered me: Can we really unlock the secrets of brain function based on the temporal resolution of current fMRI technology?" Professor Jang-Yeon Park said
.

To find a breakthrough to improve fMRI performance, we need to start
with how it works.
In fact, fMRI does not directly observe neuronal activity, but relies on another indicator to measure indirectly, which is the Blood Oxygenation Level Dependent (BOLD).

Simply put, in brain areas where nerves are active, local oxygen consumption increases and blood flow increases, which causes changes in the local magnetic field, and magnetic resonance imaging reflects neural activity
through such changes.

As mentioned earlier, the imaging capabilities of fMRI are limited
by spatial resolution as well as temporal resolution.
Among them, the main reason for the lack of spatial resolution is that traditional fMRI can detect larger blood vessels in the brain, which affect a large area of the cortex, and in turn, the activity of these cortexes will be reflected in the same blood vessels, resulting in insufficient
spatial resolution.
The solution is also clear: use a stronger magnetic field and focus on smaller blood vessels
.
At present, ultra-high-field fMRI technology has greatly improved spatial resolution
.

In contrast, the scientific community has not found a solution to the improvement of temporal resolution for a
long time.
As introduced earlier, fMRI will track changes in blood flow in the brain area where oxygen consumption suddenly increases, and this signal can be delayed by up to 1 second compared to neural activity - don't underestimate this second, you must know that nerve signals are transmitted in milliseconds, and the entire cognition, decision-making and other activities only need 0.
1 seconds
.
As a result, existing resolutions simply don't really capture neural activity
.

The latest research uses ideas for improvement that are simple and effective
.
While existing fMRIs "take" a complete image of a specific cross-section of a brain region every few seconds, Jang-Yeon Park's team has the strategy of shortening the interval to a few milliseconds — at the cost of only a small area at a time, but stitching these local images together gives a high-resolution complete image
of the cross-section of the brain.
To do this, you only need to modify some software configurations
on the basis of the existing ultra-high field fMRI.

▲DIANA can achieve direct imaging of neural activity with high spatiotemporal resolution (Image source: Reference [1]).

This study validated the feasibility
of DIANA in mouse experiments.
The team placed a anesthetized mouse in an MRI scanner with a beard pad on its face receiving a weak electrical stimulation
every 0.
2 seconds.
Between electrical stimulation, the instrument scans a tiny area of the mouse's brain every 5 milliseconds to obtain a local image, which is eight times faster
than the existing technology.

When the software integrates data from multiple scans, the process directly generates images
of mouse brain slices.
Signals from the somatosensory cortex are obtained through this method, which is the brain area
that senses the stimulation of the beard.
The resulting signal also becomes an unprecedented millisecond-level resolution fMRI image
.
According to the paper, compared to existing techniques with high latency, DIANA enables direct imaging
of neural activity.
At the same time, its spatial resolution is 0.
22 mm, which is comparable
to ultra-high-field fMRI technology.

The contemporaneous opinion article pointed out that DIANA breaks the current limitations of fMRI in terms of temporal resolution and has exciting application potential
.

▲Continuous detection of different areas enables rapid fMRI imaging (Image source: Reference [1]).

IT'S WORTH MENTIONING THAT THE RESEARCH TEAM HASN'T FULLY FIGURED OUT WHY THEIR DIANA CAN PERFORM SO WELL
.
One hypothesis is that the change in neuronal membrane potential is reflected in the lateral relaxation time of the MRI signal, that is, the speed at which the MRI signal disappears
.
Their initial hypothesis was that an increase
in signal intensity could be observed during neural activity due to prolonged lateral relaxation time due to membrane potential.

Although the biological mechanism behind DIANA's signaling is questionable, some other scientists are not worried
.
Professor Ravi Menon, a neuroscientist and physicist at Weston University in Canada, said: "The data itself shows that changes in MRI data are closely linked
to brain activity, regardless of the mechanism.
I think that's what matters in the beginning – and the details can be worked out
later.
”

Dr.
Padmavathi Sundaram, a lecturer at Massachusetts General Hospital, said she was somewhat concerned that external fluids might affect the signals, but the authors have done their best to prove that these signals are real: "This may be the first credible MRI image
of neuronal electrical currents in vivo that I have seen.
" ”

Another interesting thing is that the technique used in this study is not only simple, it is not even new
.
Similar approaches were used 20 years ago, and Professor Jang-Yeon Park's research was inspired
by previous research.
Portuguese neuroscientist Dr.
Noam Shemesh believes that previous researchers may not have been "bold" to look for signals in such a short period of time, and they may never have expected to find something
here.

"Most people think there's just noise here, and Jang-Yeon Park and others are trying to look for some signals
from that.
Apparently, they did
.
Dr.
Shemesh said
.

▲ By stimulating the mouse beard pad, the study achieved direct imaging of the neural activity of the mouse (Image source: Professor Jang-Yeon Park).

The publication of the latest Science paper represents a recognition of DIANA by the academic community, but for the technology to be truly established and applied, it still requires a lot of work
from all neuroscientists and MRI physicists.

The first thing to do, naturally, is to replicate Professor Jang-Yeon Park's results in different labs to confirm the reliability
of DIANA.
Another challenge was eventually changing the measurement from mice to humans
.
"We can't anesthetize people to get images," Dr.
Sundaram said
.
But Dr.
Shemesh believes the challenge is not insurmountable, given that MRI methods
that are more sensitive to subject movement have been dealt with before.

If the DIANA technology is eventually validated by more labs, it will open up a whole new path
in neuroscience.
Subjects can directly read neural activity in a non-invasive way by simply putting them into the instrument – a revolutionary breakthrough that will provide unprecedented tools
for understanding brain mechanisms and diseases.
We expect that this progress will be scientifically tested and ultimately benefit all humankind
.