Posts Tagged ‘ smell ’

Predictive Scents

An adorable kitty's piriform cortex suspects tuna lies nearby.

Just before we take a whiff of a smell, our brains encode a “predictive template” of the expected scent, says a new study.

This means that if we’re verbally primed with a scent-related cue – say, someone asking us whether a glass of milk has gone rotten – we’re more likely to detect a tinge of rottenness in the milk’s smell, whether it’s actually gone sour or not.

The relationship between priming and perception is a hot topic for debate in today’s neuroscience and psychology labs (I’ve written here about the influence of verbal priming on visual working memory, and here about its influence on mathematical intuition). Though psychologists have understood for decades that priming often affects what we think and experience, the neural underpinnings of this phenomenon have remained somewhat poorly understood.

But now, as the journal Neuron reports, a team led by Christina Zelano at Northwestern University’s Feinberg School of Medicine have discovered some exciting new specifics about the links between what we expect and what we smell.

The researchers monitored the brain activity of volunteers in an fMRI scanner, as the subjects tried to identify a specific smell – watermelon or Play-Doh – in a series. They found that, even before the volunteers actually smelled an odor, their brains showed activity patterns that closely matched those associated with the scent itself:

Ensemble activity patterns in anterior piriform cortex (APC) and orbitofrontal cortex (OFC) reflected the attended odor target both before and after stimulus onset. In contrast, prestimulus ensemble representations of the odor target in posterior piriform cortex (PPC) gave way to poststimulus representations of the odor itself.

In other words, the anterior piriform cortex (APC) – an area crucial for processing smells – and the orbitofrontal cortex (OFC) – an area involved in anticipating upcoming events – both lit up with activity before, during, and after the actual smell was presented. But activity patterns in the posterior piriform cortex (PPC) shifted in response to new scents – which strongly implies that this region helps compare the scents we expect against those we actually experience.

The researchers take these discoveries to mean that activity in the PPC helps shape what we smell, but doesn’t completely control it:

The robustness of target-related patterns in PPC predicted subsequent behavioral performance. Our findings directly show that the brain generates predictive templates or “search images” in PPC, with physical correspondence to odor-specific pattern representations, to augment olfactory perception.

In short, while the PPC might help us distinguish one odor from another, it doesn’t directly tell us what we smell. This fits in neatly with recent discoveries about priming in the visual pathway, which seems to function in much the same way.

As the inventor and philosopher Ray Kurzweil wrote, “We don’t actually see things – we essentially hallucinate them in detail from … low-resolution cues.”

It’s just one more confirmation that reality isn’t something we passively take in – it’s something our brains actively construct throughout each moment.

So, the next time something smells funny to you, take a moment to consider what you’re expecting to catch a whiff of. Your anticipation might be stinking up the joint more than you’d think.

Smell and Self-Control

Our sense of smell is intimately linked with our ability to make good decisions, a new study shows.

"I'm to'lly fine - jus' drag the karaoke machine over here, k?"

This might sound like a weird idea, but once you understand the underlying neuroscience, it actually makes perfect sense. It all comes down to a brain region called the orbitofrontal cortex (OFC), which is crucial for an odd grab-bag of functions: recognizing smells, anticipating rewards and punishments, and making conscious decisions.1

This new research, published in the journal PLoS ONE, focused on alcoholics in particular. A team led by Pierre Maurage at Belgium’s Catholic University of Louvain found that people with impaired executive function - the ability to make conscious decisions that take previous experiences into account – also tend to have an impaired ability to tell smells apart:

Olfactory [i.e., smell] processing can be separated into a primary “sensory” level, indexed by odour detection threshold, and a secondary “cognitive” level, indexed by odour discrimination and identification … alcoholism [probably] does not lead to a general olfactory deficit, but rather to a specific impairment for high-level olfactory processing.

In other words, the OFC isn’t involved in detecting smells, but in our ability to classify them.

As you can probably tell by now, no one’s sure what, exactly, the OFC’s role in human cognition is. This region may just deal with a wide range of processes – or it could be that our models of consciousness are missing (or mis-defining) something. It’s a genuine neuroscience mystery.

Anyway, since lots of previous studies have linked high-level olfactory impairment with diseases like schizophreniadementia, and anorexia, these new results help make a strong case that the OFC is involved these kinds of problems – and that a smell test may help doctors diagnose them more quickly:

Olfactory abilities could represent a cognitive marker of psychiatric states, and … olfactory stimulation could constitute a new tool to explore the structural and anatomical dysfunctions of the brain regions (particularly the orbitofrontal cortex) associated with high-level cognitive functions.

For this research, the team started with 20 hospital patients who had been diagnosed with alcoholism. All the patients had gone at least two weeks without having a drink – which is way longer than a lot of my friends can go.

The researchers tested the patients’ sense of smell first, by having them play one of my favorite games from preschool: Sniff the Marker. In this case, odorless felt-tip pens were filled with fluids containing various smelly chemicals (they’re called “odorants,” but saying “smelly chemicals” is more fun) diluted to various strengths. The patients took a multiple-choice test asking them to identify the smell from each marker.

Then, the researchers tested those same patients on some executive function skills, by having them play two games. In the first game, the patients had to learn a simple response to a certain cue in a series, then quickly learn to respond to a different cue instead. In the second game, they had to remember which of the pictures in a set were shown to them before the start of the round – then switch to remembering which ones hadn’t been shown before.

By comparing these test results with equivalent data from 20 healthy volunteers, the scientists discovered some interesting correlations. Alcoholic patients didn’t have a less sensitive sense of smell, but they did have much more trouble telling similar smells apart. On the games, both groups performed equally well when learning the initial skills, but the alcoholics had more trouble learning the switched-up rules – especially in the picture memory game.

This seems to point toward the idea that alcoholics have trouble consciously calling up specific memories, which might help explain why their addictive behavior tends to rely on raw feelings over logic.

Still, these results are only an early step in the road toward understanding the OFC, and its role in mental disorders. For one thing, it’d help to study alcoholics who’ve been sober for more than two weeks. It’ll also be useful to correlate these findings with fMRI data, and learn more about just what’s going on in the OFCs of these patients.

So, I think what we can all take away from this is: if you have a friend who runs up vast bar tabs and thinks Don’t Stop Believing is the greatest song of all time, get that person to sniff some scented markers, and record the results. You know, for science.


1. As I’ve mentioned before, cannabis (marijuana) seems to throw the OFC’s anticipation processing out of whack, and make the mind more likely to exaggerate the expected results of actions (both good and bad). There’s also evidence that these expectations may skew toward the negative side over time in habitual users.


Get every new post delivered to your Inbox.

Join 74 other followers