Cocktail-Party Neuroscience

"That's all very well, but what about the real world?"

This, or something to this effect, is a stock criticism of much of psychology and cognitive neuroscience. Studies of human behavior and brain function under carefully controlled laboratory conditions don't tell us much about everyday life, the argument goes.

It's a serious point. But a group of neuroscientists have now sought to dispel such worries in rather spectacular fashion. With the help of some nifty wireless headsets, Alan Gevins and colleagues of San Francisco took electroencephalography (EEG) out of the lab and organized an EEG party - allowing them to record brain electrical activity from 10 people as they chatted and drank vodka martinis. An electroencephalorgy one might say.

This is perhaps the only time in history that scientists have admitted, on record, to getting drunk with their research funding.

Pics or it didn't happen? They have pics:

And more:

The odd device held by the girl in blue is an alcohol breathalyser, which brings us onto the purpose of the study, which was to measure the effect of alcohol on brain activity.

The authors first measured the effect of alcohol on brain alpha, beta and theta band activity under standard lab conditions, and then checked to see if the results translated to the party. They did, surprisingly well in fact (although the whole thing relied on a multivariate model of the kind that make purists suspicious.)

Still, only 40% of the party data was deemed unusable due to electrical artifacts caused by participants speaking, swallowing, chewing and so forth, which is pretty good, and suggests that real-world EEG could be much more feasible than many neuroscientists would have predicted (given how annoying these sources of noise can be even under lab conditions I'd have guessed it would be more like 90%).

Now I'll make an admission: when I first read this paper, I was cynical. I felt sure it was some kind of advert for the authors' products, probably the nifty wireless EEG caps they used. "Oh very clever," I thought. "You run a wacky study, it goes viral, and you get free advertising. Well, it's worked on me, but I'm going to call you out on it."

However, the authors were one step ahead, because the paper assures readers that:

The authors are employed by the San Francisco Brain Research Institute and SAM Technology which are 100% supported by competing research grants from the U.S. Federal Government... The organization only performs research and offers no services or products. None of the authors perform consulting work. It is very unlikely that any corporation, investor, etc. would find it commercially worthwhile to buy or license the technologies that the authors have made to do their research.

OK then.

Gevins A, Chan CS, and Sam-Vargas L (2012). Towards measuring brain function on groups of people in the real world. PloS one, 7 (9) PMID: 22957099

Geometric Illusions in Astronauts

Geometric illusions in astronauts sounds like the title of a late 70s prog album, but it's actually the topic of a remarkable psychology paper just published.

Authors Gilles Clement and colleagues of the impressively-named International Space University were interested in the effects of zero gravity on optical illusions and the perception of shape.

They hypothesized that our sense of gravity pointing down (via the inner ears) is responsible for certain visual illusions. In the Inverted T illusion, for example, two lines of equal length seem different, the vertical line appearing longer than the horizontal one.

So they took 8 astronauts due to spend time on the International Space Station (ISS). Using a computer display setup, the astronauts were tested before, during, and after their time in orbit - the spaceflights lasting up to 180 days.

The results showed that the magnitude of the inverted T illusion was decreased after a long spaceflight,  although only a little bit. See above.

Also, when asked to draw a square and an even-sized cross, the astronauts tended to draw the vertical lines shorter when in space, compared to back on earth: unlike with the inverted T illusion, this happened immediately, rather than only after prolonged time in zero G.

For two other illusions, the Muller-Lyer and the Ponzo, there was no effect, however.

The authors write that these data support the idea that our sense of gravity affects our visual perception, and say that this is consistent with what happens in patients with inner-ear damage that affects the sense of up and down.

The data are pretty noisy because of the small sample size, but for once, the authors have a pretty good excuse for why it was hard to recruit participants...

Clément G, Skinner A, Richard G, and Lathan C (2012). Geometric illusions in astronauts during long-duration spaceflight. Neuroreport PMID: 22955144

When Data Filtering Introduces Bias (fMRI Edition)

A couple of months ago I blogged about a paper showing that 'filtering' of EEG data can create spurious effects.

Now, we read about another form of bias that filters can introduce, this time for fMRI: Filtering induces correlation in fMRI resting state data.

Australian neuroscientists Catherine Davey and colleagues consider temporal filtering of fMRI data in studies looking at correlation (brain functional connectivity).

Because both very high frequency and very slow changes in the fMRI signal are probably caused by artefacts, rather than interesting brain signals, it's common to use a filter to try and extract the medium-frequency changes that are of most interest (e.g. approximately 0.01 to 0.1 Hz).

However, while this filtering is very useful, Davey et al show that it can - ironically - create artefacts of its own: here's the data from one volunteer scanned during a simple task and then analyzed in 4 different ways:

Without filtering (A) there's a huge amount of 'connectivity' - too much to be realistic. This is why filtering is important.

But filtering, without correcting for the effects of the filter, actually makes things worse (B). It solves one problem but at the cost of creating another. The problem is those pesky autocorrelations. The authors say, however, that they've calculated a way to correct for filter-induced correlations (D) and that this gives more realistic results. They recommend that this should be used in future connectivity studies, but don't go into much detail regarding the question of what this means for the existing literature.

Perhaps data 'filtering' is a misleading term. It implies that all you're doing is removing the unwanted noise, leaving pristine, crystal clear data, a bit like a water filter. Mmm. What could go wrong? In fact mathematical 'filters' can put stuff into the data as well as take it out, so should we stop using that word and just call them what they are: modifications?

Davey CE, Grayden DB, Egan GF, and Johnston LA (2012). Filtering induces correlation in fMRI resting state data. NeuroImage PMID: 22939874

Naomi Wolf's "Vagina"

Naomi Wolf's "Vagina" is full of bad science about the brain - is an article I wrote for the New Statesman. It's about a new book which is... not very good.

I didn't come up with the title by the way, but I do rather like it.

See also the Neurocritic's take.

This Is Your Brain On Management

Have you ever wondered whether how the brains of managers work? New research from a group of German neuroscientists and management experts reveals all: Dissociated Neural Processing for Decisions in Managers and Non-Managers

The results were rather remarkable:

Using fMRI, the researchers found that managers' brains were less active in a number of areas, compared to the brains of non-managers, when doing the same task. By contrast, managerial brains were more active than the others only in one small area (caudate nucleus). See above.

So overall, managers had less brain activation during the task. Does that mean they have defective brains? Could this be a neurobiological explanation for the behaviour of Pointy Haired Boss and David Brent?

Not at all, say the authors. The lower activation in the brains of managers means that they were more efficient:

the managers might have found a more efficient way of sorting the presented words. This might have enabled them to faster decide for their preferred category...  Managers as expert decision-makers would seek to find a rule or heuristic on which they could base their decisions. According to previous studies, this phase of rule identification would involve the caudate nucleus

While non-managers wasted brainpower on thinking through the task with several areas of their cerebral cortex, the managers (so to speak) downsized their neurological expenditure by outsourcing the work to their caudate nucleus, an area responsible for applying a simple but effective rule.

One of the problems with these kinds of group-comparison fMRI studies is that under-activation can equally well be glossed as "deficient" or "efficient". Curiously, it usually ends up being whichever fits with the author's narrative.

That's assuming you agree that the task was about "decision making". It consisted of seeing a long series of pairs of words, one "individualistic" such as 'power' and one "collectivistic" such as 'harmony'. Participants just had to pick which word they liked best. There were no right or wrong answers. I'm not sure what kind of manager would have to do anything like that in real life. Maybe a manager of a fridge magnet poetry manufacturer?

That's also assuming the results are solid. The authors provide few details on the fMRI methods (the main results are said to be "cluster-level corrected at p less than 0.0013", which is an unusual threshold to use and an extremely strict one (0.05 cluster-level is more common; this is about 40 times stricter).

Still. If you do buy these results, the message is: management is literally about using as little of your brain as possible...

Caspers S, Heim S, Lucas MG, Stephan E, Fischer L, Amunts K, and Zilles K (2012). Dissociated neural processing for decisions in managers and non-managers. PloS one, 7 (8) PMID: 22927984