Posts Tagged ‘fMRI’

Social cognitive deficits in autism spectrum disorder

May 6, 2011 3 comments One of the hallmarks of Autism Spectrum Disorder (ASD) is an impairment in social cognitive skills. This manifests in individuals with ADS having trouble orienting their attention towards people. Accordingly, they also show deficits orienting their attention in response to social cues from others, such as eye gaze, head turns and pointing gestures.

Understanding the social cognitive impairments associated with ASD has been challenging in that studies set in naturalistic settings often reveal the deficit but lab experiments performed on computers don’t.

For example, some naturalistic studies have looked at home movies of infants and found that those later diagnosed with ASD showed less social orienting and were less responsive to cues from others to orient to objects. For example, if their mom was in the room, they would look at her a lot less and they’d also be less likely to respond when their mothers tried to direct their attention to a toy in the room by looking or pointing at it.

However, people with ASD have been shown to respond to non-naturalistic social cues in the lab. Social orienting has been frequently been tested by use of a variation on Michael Posner’s spatial cueing paradigm. This works as follows:

1. Participants are seated in front of a computer
2. A stimulus – a pair of eyes gazing to either side (or straight ahead) or arrows pointing to either side or neither – appears on the screen
3. Shortly after, a stimulus (the target object) appears to one side or the other, either on the side which the eyes or arrows were pointing towards or the opposite side.
4. Participants have to indicate which side the target object appeared on by pressing either a right or left button.
5. Performance on the task is assessed by measuring the amount of time it takes to participants to press the button indicating on which side the target appeared. Most participants, including ASD patients, are as quick with the gaze cue (the eyes) as with the arrow cue.

Posner cue paradigm

(The left side of the above figure shows a single trial (with “directional eyes”), in which participants first see a fixation cross, then one of four directional/non-directional stimuli, after which the target appears either on the same side indicated by the cue or the opposite side. Participants need to indicate which side a target stimulus appeared on by pushing a button. The right side shows the three other trial types (from top to bottom): neutral arrow, directional arrow, neutral eyes)

Past studies have shown that people orient faster to cued (like in the left side of the above figure) versus noncued locations, known as the facilitation effect. Previous studies using this task have produced inconsistent results, but most of them have shown ASD populations performing comparably to non-ASD populations.

In this study, researchers used the above-described cue task to examine the neural mechanisms underlying social orienting in ASD, with the hope that if there were no behavioral differences, neural activity might reveal that ASD individuals are performing the task differently. Other studies have shown that non-ASD populations treat social and non-social cue stimuli differently. It was hoped that neural activity revealed in this study would shed light on the discrepancies in behavioral results for ASD populations in lab versus computer settings.

In terms of behavior, both the control and the ASD group showed quicker responses for gaze and arrow cues with no between group difference, which is consistent with previous lab studies.

However, neural activation patterns showed significant group differences. The control group showed greater activation for social vs. nonsocial cues in many different brain regions, with gaze (eyeball) cues eliciting increased activity in many frontoparietal areas, supporting the idea that neurotypical brains treat social stimuli different from non-social stimuli. The ASD group, on the other hand, showed much less difference in neural activation between social vs. non-social cues. Although these differences in neural activation are too numerous to cover here, one region of interest, superior temporal sulcus (STS), stood out. The STS has been shown to be associated with the perception of eye gaze and other work has suggested the region may be involved in understanding the intentions and mental states of others. In this study, ASD individuals showed decreased STS in the gaze cue condition (versus controls). This data suggests that the STS may not be sensitive toward the social significance of eye gaze in ASD individuals.

The authors point out that although ASD individuals don’t seem to rely on the same neural circuitry to perceive social cues such as eye gaze, they have found a way to use the low-level perceptual information available in social cues to adapt a strategy that allows them to discern that gaze direction conveys meaning about the environment. That being said, ASD individuals mostly don’t do this very well in more naturalistic environments. So, although this strategy might work in a scanner with “cartoon” eyes and where there are no environmental distractions, it’s unlikely that ASD individuals could adapt this strategy in a naturalistic environment. On the contrary, one could also frame these results from the perspective of the ASD individual; that is, given the non-naturalistic environment of the scanner, and the fact that the task demands were very simple and not dependent on social cognitive processing, why should non-ASD individuals treat the gaze vs. arrow stimuli differently? Why not just rely on low-level information and thus expend less cognitive energy? It’s a good example of the automaticity of social cognitive processes. Give humans a set of cartoon eyeballs to look at and they can’t help but process these as distinct from something non-social.

An additional take away from this paper is that even when one finds no behavioral differences between groups, there might be some interesting differences in neural activity worth exploring via fMRI or EEG.


Greene DJ, Colich N, Iacoboni M, Zaidel E, Bookheimer SY, & Dapretto M (2011). Atypical neural networks for social orienting in autism spectrum disorders. NeuroImage, 56 (1), 354-62 PMID: 21334443

The Neural Correlates of Romantic Love

May 1, 2011 Leave a comment

For the most part, fMRI studies attempt to localize cognitive processes to specific regions in the brain. Popular media often introduce these studies with headlines that tout the discovery of “the brain region” for memory, language, empathy, moral reasoning, loving weiner schnitzel and so on.

These headlines can be terribly misleading, as they’re often misinterpreted to suggest a specific brain region is dedicated to a single function, when, in fact, any given function maps on to a network of regions (forming a circuit), while any given region is part of multiple circuits subserving many functions. Similar faux pas can be found in descriptions of the functions associated with genes, e.g. “The gene for (fill in the blank).”

A few years back, the NY Times ran an infamous piece featuring the work of a neuromarketing company. In a horrible experiment fit for The Onion, participants lay in the scanner while looking at pictures of then presidential candidates. Subjects showed increased amygdala activation to pictures of Mitt Romney, which researchers interpreted as a sign of anxiety.

But after watching Romney speak on video, the amygdala activity died down, which researchers said showed that voters’ anxiety had decreased.

Meanwhile subjects’ anterior cingulates lit up to pictures of Hillary Clinton.

Here’s how researchers interpreted this neural activity:

Emotions about Hillary Clinton are mixed. Voters who rated Mrs. Clinton unfavorably on their questionnaire appeared not entirely comfortable with their assessment. When viewing images of her, these voters exhibited significant activity in the anterior cingulate cortex, an emotional center of the brain that is aroused when a person feels compelled to act in two different ways but must choose one. It looked as if they were battling unacknowledged impulses to like Mrs. Clinton.

The Times article about the “research” was quickly and roundly criticized by prominent neuroscientists, 17 of whom quickly responded with a signed letter to the editor, which the Times ran a couple of days later:

To the Editor:

“This Is Your Brain on Politics” (Op-Ed, Nov. 11) used the results of a brain imaging study to draw conclusions about the current state of the American electorate. The article claimed that it is possible to directly read the minds of potential voters by looking at their brain activity while they viewed presidential candidates.

For example, activity in the amygdala in response to viewing one candidate was argued to reflect “anxiety” about the candidate, whereas activity in other areas was argued to indicate “feeling connected.” While such reasoning appears compelling on its face, it is scientifically unfounded.

As cognitive neuroscientists who use the same brain imaging technology, we know that it is not possible to definitively determine whether a person is anxious or feeling connected simply by looking at activity in a particular brain region. This is so because brain regions are typically engaged by many mental states, and thus a one-to-one mapping between a brain region and a mental state is not possible.As cognitive neuroscientists, we are very excited about the potential use of brain imaging techniques to better understand the psychology of political decisions. But we are distressed by the publication of research in the press that has not undergone peer review, and that uses flawed reasoning to draw unfounded conclusions about topics as important as the presidential election.

Adam Aron, Ph.D., University of California, San Diego
David Badre, Ph.D., Brown University
Matthew Brett, M.D., University of Cambridge
John Cacioppo, Ph.D., University of Chicago
Chris Chambers, Ph.D., University College London
Roshan Cools, Ph.D., Radboud University, Netherlands
Steve Engel, Ph.D., University of Minnesota
Mark D’Esposito, M.D., University of California, Berkeley
Chris Frith, Ph.D., University College London
Eddie Harmon-Jones, Ph.D., Texas A&M University
John Jonides, Ph.D., University of Michigan
Brian Knutson, Ph.D., Stanford University
Liz Phelps, Ph.D., New York University
Russell Poldrack, Ph.D., University of California, Los Angeles
Tor Wager, Ph.D., Columbia University
Anthony Wagner, Ph.D., Stanford University
Piotr Winkielman, Ph.D., University of California, San Diego

Undoubtedly, fewer people saw that letter than saw the original article, which was much more prominently displayed.

(By the above study’s logic, looking at a picture of Donald Trump should elicit activity in the anterior insula, a region often associated with disgust responses)

It’s unfortunate that the study received such a prominent platform for distribution because people, especially non scientists, can be heavily influenced by articles with pictures of brains or technical sounding neuro language. One study, which I’ve written about on the blog, found that people were much more likely to believe a nonsensical article if it had meaningless neuroscience language in it than if it didn’t. As the average lay person doesn’t possess the technical skills to distinguish between valid and invalid fMRI studies, it’s up to the scientific community to police itself, which it does a pretty good job of through the peer review process.

This next study I’ll talk about demonstrates some of the challenges inherent to mapping localized neural activity onto unseen mental processes; in this case, the subjective experience of intense, romantic long-term love.

Aaron and colleagues previously published a study that presented neural correlates of intense romantic love (2005). In brief, the study reported that regions in the reward circuit of the brain were activated in response to pictures of a lover (versus a close friend). In the current study, they wanted to explore if these findings could be extended to long-term married couples (couples together for more than 20+ years who report still being madly in love).

Participants lay in the scanner and were repeatedly presented with pictures from four different categories: their partner, a close friend, and both a highly familiar and low-familiar neutral acquaintance. They were instructed to think about “experiences with each stimulus person (that were) nonsexual in nature.”

The fMRI data was analyzed via subtraction, a common fMRI analysis method in which one condition is compared to another to see if any differences fall out. The contrast of interest was between the partner and the close friend. In a cognitive process sense, the only difference thought to exist between perceiving these two individuals was thought to be the subject’s romantic love for one and not for the other. So if neural activity in the close friend condition is subtracted from activity in the partner condition, whatever is left over should represent the neural substrate of romantic love.

Researchers found activation in the ventral tegmental area, substantia nigra and nucleus accumbens (and the hippocampus, which corresponded with reported sex frequency, but the effect seems to be driven largely by two outliers, one of whom reported they have sex almost every day).

The activity does suggest a classic reward response and replicates previous findings. However, the big question isn’t whether there is a response, but rather what’s driving it?

A valid fMRI study doesn’t only rely on the integrity or analysis of the fMRI data, but, also, and perhaps more importantly, on the experimental design. In order to attribute increased activation in one condition versus the other to a specific cognitive function, one must be confident they have created conditions that have cleanly isolated the independent variable of interest (romantic love). The ventral tegmental area and other regions in the basal ganglia have been repeatedly shown to encode reward value – that is, they respond to things that give us hedonic pleasure, such as food, drugs, sex or receiving money. Past work has shown that intense romantic love is associated with activity in the those regions (Aaron 2005). In the current study, activity in some neural regions previously associated with maternal pair bonding was shown (substantia nigra, for one). The authors hypothesized that neural correlates for romantic long-term love should encompass those associated with both intense romantic love and maternal pair bonding.

But this analysis is dependent on long-term, romantic love being the only difference between conditions that would explain the differences in brain activity. And that may not entirely be the case.

Alternative Explanations
Just to refresh, the major dependent measure of interest was neural activity, especially of reward circuitry, while subjects looked at pictures of their long-term partner versus a close friend. One additional difference between these conditions (beyond romantic love) is that romantic partners are probably more familiar and closer to participants than close friends. This is a shortcoming acknowledged by the authors.

This difference suggests a causal chain of cognitive operations that could offer alternative explanations for some of the data seen in this study. First, It’s been shown that we prefer things that are familiar to us (the “mere exposure” effect, Zajonc, 1968). Second, we’re able to process familiar things (or people) much more fluently compared to the less familiar (Reber). Third, fluency processing has been associated with judgments of aesthetic appreciation such that the more fluently we can process something, the more beautiful or attractive we’re likely to rate it (Alter). Although the objective attractiveness of the photos was controlled for via a group of independent raters, the participants were likely much more subjective in their judgments and perhaps found their partners more attractive than an objective viewer might. Viewing attractive faces has been shown to elicit strong neural activity, particularly in the reward circuitry (NaCC and OFC).

Furthermore, it has been posited that people incorporate close others into their psychological construct of self. Recent studies (deGreck 2008) showed that regions active in a reward task, such as the bilateral ventral striatum, and the ventral tegmental area (VTA), are also involved in differentiating between high and low personal relevance.It seems that we find thinking about ourselves pretty damn rewarding! (We’re all at least a little bit narcissistic). To the extent that someone has been incorporated into our self concept, thinking about that person, or looking at their picture as in this study, could be correlated with responses in reward regions of the brain in part because they activate thoughts of ourselves.

Both the familiarity –> processing fluency –> attractiveness model and self relevant thinking are plausible alternative explanations for at least some of the neural correlates found in this paper.

One other potential area of concern is that there is no way of knowing that participants weren’t thinking about sex with their partners, even though they were told not to. This might be especially difficult to achieve, especially for the two outliers who reported almost daily sex. Regions active during sexual arousal include R. amygdala, hypothalamus, hippocampus, midbrain, mOFC and nucleus accumbens, many of which were found to be active in this study.

The neural activity measured here may very well reflect some aspect of individuals’ love for their partners. But there seem to be other possible explanations for some of the data. I suppose that’s why people call the study of consciousness, of which subjective experiences such as romantic love are a subset, the “hard problem.”

Acevedo BP, Aron A, Fisher HE, & Brown LL (2011). Neural correlates of long-term intense romantic love. Social cognitive and affective neuroscience PMID: 21208991

Aron, A., Fisher, H., Mashek, D., Strong, G., Li, H., Brown, L. (2005). Reward, motivation and emotion systems associated with early-stage intense romantic love. Journal of Neurophysiology, 93, 327–37.

DEGRECK, M., ROTTE, M., PAUS, R., MORITZ, D., THIEMANN, R., PROESCH, U., BRUER, U., MOERTH, S., TEMPELMANN, C., & BOGERTS, B. (2008). Is our self based on reward? Self-relatedness recruits neural activity in the reward system NeuroImage, 39 (4), 2066-2075 DOI: 10.1016/j.neuroimage.2007.11.006

Alter, A., & Oppenheimer, D. (2008). Easy on the mind, easy on the wallet: The roles of familiarity and processing fluency in valuation judgments Psychonomic Bulletin & Review, 15 (5), 985-990 DOI: 10.3758/PBR.15.5.985

Peskin, M., & Newell, F. (2004). Familiarity breeds attraction: Effects of exposure on the attractiveness of typical and distinctive faces Perception, 33 (2), 147-157 DOI: 10.1068/p5028

Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing Fluency and Aesthetic Pleasure: Is Beauty in the Perceiver’s Processing Experience? Personality and Social Psychology Review, 8 (4), 364-382 DOI: 10.1207/s15327957pspr0804_3

Regard thyself and put down the smoke stick

April 15, 2011 Leave a comment

As many a former smoker will probably attest, quitting cigarettes ranks high in the hard-to-kick category. I made several unsuccessful attempts before finally kicking the habit after a 10 year pack-a-day run. Ultimately what worked for me was to go cold turkey, but there were perhaps other alternatives which I might have tried. In a paper from Nature Neuroscience, researchers from University of Michigan provided participants with interventions involving individually tailored messages* designed to encourage quitting and found that participants’ brain activity while listening to the messages predicted how likely they would be to successfully quit smoking.

*Tailored messages are statements about an individuals’ issues and thoughts about quitting smoking, derived from pre-screen interviews with them. e.g., “You are worried that when angry or frustrated, you may light up”.

Here’s the premise: Anti-smoking messages custom made for an individual can be more effective than generic ones, but only if said individual processes those messages in a self directed manner. Past research has shown a specific set of neural regions – primarily the mPFC and precuneus/posterior cingulate – to be associated with self referential thinking. Therefore, researchers hypothesized, activity in these brain regions while processing tailored anti-smoking messages might predict the likelihood of quitting.

The Study
The experiment was carried out over three days with a follow-up visit four months later.

Day 1: 91 participants completed a health assessment, demographic questionnaire and a psychosocial characteristics scale related to quitting smoking. Responses were then used to create smoking cessation messages tailored to each individual.
Day 2: Participants went into scanner and performed 2 fMRI tasks: The first task had participants listen to anti-smoking messages of three different types: personally tailored anti-smoking, non-tailored anti-smoking and neutral.

Here are some examples of what they heard:

Tailored messages
A concern you have is being tempted to smoke when around other smokers.
Something else that you feel will tempt you after you quit is because of a craving.
You are worried that when angry or frustrated, you may light up.
Untailored messages
Some people are tempted to smoke to control their weight or hunger.
Smokers also light up when they need to concentrate.
Certain moods or feelings, places, and things you do can make you want to smoke.
Neutral messages
Oil was formed from the remains of animals and plants that lived millions of years ago.
Sighted in the Pacific Ocean, the world’s tallest sea wave was 112 feet.
Wind is simple air in motion. It is caused by the uneven heating of the earth’s surface by the sun.

Then, participants completed a self appraisal task to identify brain regions active during self relevant thought processes. In this task, participants saw adjectives appear on the screen and had to either rate how much the adjective described them or whether the adjective was positive or negative.

Day 3: Participants completed a web-based smoking cessation program and were instructed to quit smoking. (They were given a supply of nicotine patches to get themselves started)

Experimenters checked in with subjects four months later to see if they were abstaining from smoking. Out of 87 who participated in the smoking cessation program, 45 were not smoking, while 42 were still (or had quit briefly and restarted) smoking.

Subjects were given a surprise memory test for the anti-smoking messages they’d received four months prior and remembered self relevant, tailored messages most well. However, their memory performance was not related to whether they successfully quit smoking.

As for the fMRI data, experimenters used a mask of tailored vs. untailored message conditions AND self-appraisal to identify the region common to both processes. This seems like a mild case of double dipping, no? That is, finding a brain region that responds to the condition of interest (in this case, voxels more active in tailored vs. untailored conditions) and then using the same data to test the hypothesis. Ideally, the ROI would be obtained independently of the main task.

A blow by blow on the different contrasts of interest:

1. Researchers looked at brain regions more active during tailored vs. untailored messages and found differential activation in the regions below.

There are, I think, some problems here; mainly, that the task differences for processing tailored vs non-tailored statements may extend beyond self relevant thinking to (1) memory processes employed in processing either category of stimuli; that is, episodic (tailored) vs. semantic (non-tailored), (2) cognitive effort, (3) elicitation of visual vs. non-visual memory, (4) processing fluency and (5) affect or reward responses. Thus, the difference in brain activation found in this task might reflect something other than just self referential processing.

2. The localizer task (used to isolate neural areas involved in self appraisal) had participants process adjectives either by relating them to self or by judging their affective value. This suggests an alternative explanation for the categorical contrast in that it isn’t specific to self per se, but really more specific to people vs. non-people. A more widely used version of this task has participants process adjectives with regards to self or an other. As a further control, a third condition is often included in which participants identify whether words are in upper case or lower case. The contrast applied is (self – control) – (other – control). It’s not clear why the researchers chose the task they did, which seems significantly noisier.

Here’s the contrast from the present study:

And here’s a contrast from another study (Jenkins 2010) that looked at three different types of self-referential processing.

Although roughly similar, the current study shows cortical midline activation seems to be much more dorsal than that found in Jenkins (2010). Using an ROI derived from this localizer task to correlate neural activity in tailored vs. untailored statements with quitting led to a non-significant result (from supplementary materials). This could explain why the researchers used the composite mask to define the ROI.

3. Again, the primary ROI was defined as a composite of overlapping regions between the self reference task AND the tailored vs. untailored statements task, which was used to compare neural activity with quitting behavior. They found that activity in these regions – which included dmPFC, precuneus and angular gyrus – during tailored smoking cessation messages predicted the likelihood of successfully abstaining from smoking. dmPFC and precuneus activation also individually predicted smoking cessation success, although angular gyrus did not.

This study provides clear evidence that participants processed tailored vs. non-tailored messages about smoking differently, and that this difference corresponded to their ability to stop smoking. However,
(1) neither task effectively isolates self referential processing,
(2) the region of activation was much more dorsal than that usually found in this literature (Northoff & Bermpohl, 2004; Schneider et al.,2008; Uddin, Iacoboni, Lange, & Keenan, 2007; Gillihan & Farah, 2005),
(3) an independently obtained ROI yielded insignificant results and
(4) mPFC and precuneus subserve an untold number of cognitive processes beyond self reflection.

Therefore, it seems a bit of a stretch to claim the neural activation found in this study is indicative of self referential processing.

Chua HF, Ho SS, Jasinska AJ, Polk TA, Welsh RC, Liberzon I, & Strecher VJ (2011). Self-related neural response to tailored smoking-cessation messages predicts quitting. Nature neuroscience, 14 (4), 426-7 PMID: 21358641

Jenkins AC, & Mitchell JP (2010). Medial prefrontal cortex subserves diverse forms of self-reflection. Social neuroscience, 1-8 PMID: 20711940

Northoff, G. (2005). Emotional-cognitive integration, the self, and cortical midline structures Behavioral and Brain Sciences, 28 (02) DOI: 10.1017/S0140525X05400047

Gillihan, S., & Farah, M. (2005). Is Self Special? A Critical Review of Evidence From Experimental Psychology and Cognitive Neuroscience. Psychological Bulletin, 131 (1), 76-97 DOI: 10.1037/0033-2909.131.1.76

SCHNEIDER, F., BERMPOHL, F., HEINZEL, A., ROTTE, M., WALTER, M., TEMPELMANN, C., WIEBKING, C., DOBROWOLNY, H., HEINZE, H., & NORTHOFF, G. (2008). The resting brain and our self: Self-relatedness modulates resting state neural activity in cortical midline structures Neuroscience, 157 (1), 120-131 DOI: 10.1016/j.neuroscience.2008.08.014

UDDIN, L., IACOBONI, M., LANGE, C., & KEENAN, J. (2007). The self and social cognition: the role of cortical midline structures and mirror neurons Trends in Cognitive Sciences, 11 (4), 153-157 DOI: 10.1016/j.tics.2007.01.001

fMRI lie detection takes a hit

June 23, 2010 Leave a comment

Companies like No Lie fMRi and Cephos took a hit with a recent ruling that fMRI would not be allowed as evidence of lying in a criminal court case .

Categories: fMRI, Human, neuroscience Tags: ,

Get every new post delivered to your Inbox.

%d bloggers like this: