neuroscience

To Pay Attention, the Brain Uses Filters, Not a Spotlight

A brain circuit that suppresses distracting sensory information holds important clues about attention and other cognitive processes.
The illustration shows ghostly hands partially obstructing a person’s sight and hearing.

We attend to only a fraction of the sensory data available to us. New results are helping to explain how the brain filters out the sensations least interesting to it at any moment.

Jason Lyon for Quanta Magazine

We can pick out a conversation in a loud room, amid the rise and fall of other voices or the hum of an air conditioner. We can spot a set of keys in a sea of clutter, or register a raccoon darting into the path of our onrushing car. Somehow, even with massive amounts of information flooding our senses, we’re able to focus on what’s important and act on it.

Attentional processes are the brain’s way of shining a searchlight on relevant stimuli and filtering out the rest. Neuroscientists want to determine the circuits that aim and power that searchlight. For decades, their studies have revolved around the cortex, the folded structure on the outside of the brain commonly associated with intelligence and higher-order cognition. It’s become clear that activity in the cortex boosts sensory processing to enhance features of interest.

But now, some researchers are trying a different approach, studying how the brain suppresses information rather than how it augments it. Perhaps more importantly, they’ve found that this process involves more ancient regions much deeper in the brain — regions not often considered when it comes to attention.

By doing so, scientists have also inadvertently started to take baby steps toward a better understanding of how body and mind — through automatic sensory experiences, physical movements and higher-level consciousness — are deeply and inextricably intertwined.

Hunting for Circuits

For a long time, because attention seemed so intricately tied up with consciousness and other complex functions, scientists assumed that it was first and foremost a cortical phenomenon. A major departure from that line of thinking came in 1984, when Francis Crick, known for his work on the structure of DNA, proposed that the attentional searchlight was controlled by a region deep in the brain called the thalamus, parts of which receive input from sensory domains and feed information to the cortex. He developed a theory in which the sensory thalamus acted not just as a relay station, but also as a gatekeeper — not just a bridge, but a sieve — staunching some of the flow of data to establish a certain level of focus.

But decades passed, and attempts to identify an actual mechanism proved less than fruitful — not least because of how enormously difficult it is to establish methods for studying attention in lab animals.

That didn’t stop Michael Halassa, a neuroscientist at the McGovern Institute for Brain Research at the Massachusetts Institute of Technology. He wanted to determine exactly how sensory inputs got filtered before information reached the cortex, to pin down the precise circuit that Crick’s work implied would be there.

He was drawn to a thin layer of inhibitory neurons called the thalamic reticular nucleus (TRN), which wraps around the rest of the thalamus like a shell. By the time Halassa was a postdoctoral researcher, he had already found a coarse level of gating in that brain area: The TRN seemed to let sensory inputs through when an animal was awake and attentive to something in its environment, but it suppressed them when the animal was asleep.

In 2015, Halassa and his colleagues discovered another, finer level of gating that further implicated the TRN as part of Crick’s long-sought circuit — this time involving how animals select what to focus on when their attention is divided among different senses. In the study, the researchers used mice trained to run as directed by flashing lights and sweeping audio tones. They then simultaneously presented the animals with conflicting commands from the lights and tones, but also cued them about which signal to disregard. The mice’s responses showed how effectively they were focusing their attention. Throughout the task, the researchers used well-established techniques to shut off activity in various brain regions to see what interfered with the animals’ performance.

As expected, the prefrontal cortex, which issues high-level commands to other parts of the brain, was crucial. But the team also observed that if a trial required the mice to attend to vision, turning on neurons in the visual TRN interfered with their performance. And when those neurons were silenced, the mice had more difficulty paying attention to sound. In effect, the network was turning the knobs on inhibitory processes, not excitatory ones, with the TRN inhibiting information that the prefrontal cortex deemed distracting. If the mouse needed to prioritize auditory information, the prefrontal cortex told the visual TRN to increase its activity to suppress the visual thalamus — stripping away irrelevant visual data.

The attentional searchlight metaphor was backward: The brain wasn’t brightening the light on stimuli of interest; it was lowering the lights on everything else.

Despite the success of the study, the researchers recognized a problem. They had confirmed Crick’s hunch: The prefrontal cortex controls a filter on incoming sensory information in the thalamus. But the prefrontal cortex doesn’t have any direct connections to the sensory portions of the TRN. Some part of the circuit was missing.

Until now. Halassa and his colleagues have finally put the rest of the pieces in place, and the results reveal much about how we should be approaching the study of attention.

Obscuring, Dimming, Blinking

With tasks similar to those they’d used in 2015, the team probed the functional effects of various brain regions on one another, as well as the neuronal connections between them. The full circuit, they found, goes from the prefrontal cortex to a much deeper structure called the basal ganglia (often associated with motor control and a host of other functions), then to the TRN and the thalamus, before finally going back up to higher cortical regions. So, for instance, as visual information passes from the eye to the visual thalamus, it can get intercepted almost immediately if it’s not relevant to the given task. The basal ganglia can step in and activate the visual TRN to screen out the extraneous stimuli, in keeping with the prefrontal cortex’s directive.

“It’s an interesting feedback pathway, which I don’t think has been described before,” said Richard Krauzlis, a neuroscientist at the National Eye Institute at the National Institutes of Health in Maryland who did not participate in this study.

Furthermore, the researchers found that the mechanism doesn’t just filter out one sense to raise awareness of another: It filters information within a single sense too. When the mice were cued to pay attention to certain sounds, the TRN helped to suppress irrelevant background noise within the auditory signal. The effects on sensory processing “can be much more precise than just suppressing the whole thalamic region for one sensory modality, which is a rather blunt form of suppression,” said Duje Tadin, a neuroscientist at the University of Rochester.

“We often neglect how we get rid of the things that are less important,” he added. “And oftentimes, I think that’s a more efficient way of dealing with information.” If you’re in a noisy room, you can try raising your voice to be heard — or you can try to eliminate the source of the noise. (Tadin studies this kind of background suppression in other processes that happen more quickly and automatically than selective attention.)

Halassa’s findings indicate that the brain casts extraneous perceptions aside earlier than expected. “What’s interesting,” said Ian Fiebelkorn, a cognitive neuroscientist at Princeton University, is that “filtering is starting at that very first step, before the information even reaches the visual cortex.”

Counting how many times the team in white passes the basketball reveals one surprising aspect of how selective attention works.

Counting how many times the team in white passes the basketball reveals one surprising aspect of how selective attention works.

Transport for London

There’s an obvious weakness in the brain’s strategy of tossing out sensory information this way, though — namely, the danger that the jettisoned perceptions might be unexpectedly important. Work by Fiebelkorn suggests that the brain has a way to hedge against those risks.

When people think about the searchlight of attention, Fiebelkorn says, they think of it as a steady, continuously shining beam that illuminates where an animal should direct its cognitive resources. But “what my research shows is that that’s not true,” he said. “Instead it seems that the spotlight is blinking.”

According to his findings, the focus of the attentional spotlight seems to get relatively weaker about four times per second, presumably to prevent animals from staying overly focused on a single location or stimulus in their environment. That very brief suppression of what’s important gives other, peripheral stimuli an indirect boost, creating an opportunity for the brain to shift its attention to something else if necessary. “The brain seems to be wired to be periodically distractible,” he said.

Fiebelkorn and his colleagues, like Halassa’s team, are also looking to subcortical regions to explain this wiring. For now, they’ve been studying the role of yet another section of the thalamus, but they plan to look into the basal ganglia in the future, too.

Grounding Cognition in Action

These studies mark a critical shift: Attentional processes were once understood to be the province of the cortex alone. But according to Krauzlis, over the past five years “it’s become a little more obvious that there are things that are happening underneath the cortex.”

“Most people want the cerebral cortex to do all the heavy lifting for us, and I don’t think that’s realistic,” said John Maunsell, a neurobiologist at the University of Chicago.

Correction added on Sept. 23: Michael Halassa’s affiliation with the McGovern Institute for Brain Research was added.

Portrait photograph of Michael Halassa.

Halassa gives his attention to a multi-electrode array, a device that he and his colleagues used to control brain activity during their search for the neural circuitry that filters out distracting stimuli.

Caitlin Cunningham Photography

In fact, Halassa’s discovery of the basal ganglia’s role in attention is particularly fascinating. That’s partly because it is such an ancient area of the brain, one that hasn’t typically been viewed as a part of selective attention. “Fish have this,” Krauzlis said. “Going back to the earliest vertebrates, like the lamprey, which doesn’t have a jaw” — or a neocortex, for that matter — “they have basically a simple form of basal ganglia and some of these same circuits.” The fishes’ neural circuitry may offer hints about how attention evolved.

Halassa is particularly intrigued by what the connection between attention and the basal ganglia might reveal about conditions like attention deficit hyperactivity disorder and autism, which often manifest as hypersensitivity to certain kinds of inputs.

But perhaps the most profoundly interesting point about the involvement of the basal ganglia is that the structure is usually associated with motor control, although research has increasingly implicated it in reward-based learning, decision-making and other motivation-based types of behavior as well.

With the work being done in Halassa’s lab, the basal ganglia’s role has now been extended to include sensory control too. This highlights the fact that “attention is really about sequencing from this to that in the correct order and making sure you don’t get distracted by things you shouldn’t be distracted by,” Maunsell said. “The notion that motor structures are involved in this … is appropriate, in a way — that they should be right at the heart of the process of deciding what you will attend to next, what you will focus your sensory resources on next.”

That’s in keeping with a burgeoning view of attention — and cognition as a whole — as processes based on what’s known as active inference. The brain doesn’t passively sample information from the environment and then respond to the observed external stimuli. The reverse also happens, with body movements as small as the flicker of an eye also guiding perception. The sensory and motor systems “don’t operate independently, and they evolved together,” Fiebelkorn said. And so motor regions don’t only help to shape the output (an animal’s behavior); they also help to shape the input. Halassa’s findings provide further support for that more proactive role.

“Perception serves action, because we have to represent the world in order to act in it,” said Heleen Slagter, a cognitive scientist at VU University Amsterdam. “How we learn to perceive the world around us is very much through action.” The high level of interconnection with the cortex suggests that, even beyond attention, “these subcortical structures play a much more important role in higher-order cognition than I think is often considered.”

And that, in turn, could provide hints about how to think about consciousness, neuroscience’s most elusive subject. As evidenced by Halassa’s study and other research, “when we look at the neural correlates of attention, we’re actually looking to some extent at the neural correlates of perception,” Maunsell said. “It’s part of a bigger picture, in terms of trying to understand how the brain works.”

Slagter is now studying the role that the basal ganglia might play in consciousness. “We experience the world not just using our bodies, but because of our bodies. And brains represent the world in order to meaningfully act in it,” she said. “Therefore, I would think that conscious experience must be tightly linked to actions,” just like attention. “Consciousness should be action oriented.”

This article was reprinted on TheAtlantic.com.

Comment on this article