Alpha activity in the EEG is associated with attention and internal imagery but questions remain.
Alpha oscillations are perhaps the most noticeable feature of EEG signals, dominant when you have your eyes closed and popping up at times when you have your eyes open too. In an earlier post we showed data demonstrating a large variation in oscillation activity across human populations. What is their purpose?
One theory suggests that alpha oscillations have something to do with selective attention. Attention is your brain’s mechanism for focusing on what is important to you at each moment in time, whilst suppressing any irrelevant and potentially distracting information – a kind of signal-to-noise optimization of your sensory and internal cognitive environment.
Theories of Alpha Oscillations
Because of the large and visible increase in alpha oscillations when you close your eyes, it was initially thought that these oscillations represented “cortical idling”. However, over the last few decades, scientists have shown a connection between alpha oscillations and attention, turning this theory on its head
One of the most commonly cited theories was proposed by Wolfgang Klimesch who, in a similar vein to proposals from Ole Jensen, suggests that increases in synchronisation of alpha oscillations (resulting in higher amplitude/power) are involved in the suppression of information which is not relevant to the task at hand, whilst increases in the desynchronisation of alpha oscillations releases this inhibition to facilitate task relevant processing.
There are however, conflicting theories on how it works (for instance see a review here). One of the difficulties and confounds of interpretation arises due to the fact that scientists often use methods that measure the overall power or amplitude of activity in the alpha band in a manner that does not distinguish the oscillatory or periodic component from the alpha that forms a normal part of the 1/f spectrum (see here for more discussion). Nonetheless, here is a review of the some of the findings.
Spatial Cuing studies
Many studies have explored the role of alpha in attention using spatial attention cuing tasks, similar to those first proposed by Michael Posner in the 1980s. They have showed, using a spectral analysis technique called “temporal-spectral-evolution” (TSE), that after being presented with a spatial cue (either visual or auditory), indicating where a target is about to appear, you get an increase in the amplitude of activity in the alpha band. This increase is observed across posterior parietal and occipital regions in the hemisphere of the brain, which represents the visual field where the target is NOT going to appear (which is actually ipsilateral to – the same side as – the target location because of hemispheric cross over of visual processing pathways in the brain). In some studies this is shown as a contralateral decrease in alpha, rather than an ipsilateral increase. Studies by Michael Worden, Gregor Thut and Kai-Ming Fu have shown this change in alpha between an unattended location versus an attended location to be in the range of 7-22%. Alpha increases in the context of tasks like this have been proposed to reflect the inhibition, or gating, of the incoming visual signals arising from the uncued, and therefore irrelevant, location. However, the reporting of contralateral decreases in alpha suggests that alpha may also facilitate attentional mechanisms within the target location as an enhancement of task relevant signals, not just through the inhibition of task-irrelevant information – a theory supported by Satu and Matias Palva. However, one major caveat to these studies is that the TSE method doesn’t differentiate between whether the alpha is simply a result of changing complexity of the temporal patterns, or whether it is specifically reflecting oscillations per se.
Internal Attention or Mental Imagery
Other studies have focused more on how alpha is related to internal attention – the attention that is required to perform mental imagery, or manipulate and retain information in working memory.
For example, a study by Nicolas Cooper analyzed alpha activity when people either had to pay attention to sensory cues (visual, acoustic or haptic) or perform a mental imagery task in the equivalent sensory domain (with their eyes open). They showed that alpha activity across occipital regions were greater when people performed mental imagery, compared to when they were paying attention to sensory stimuli. This effect was largest for the visual modality (41% increase for visual compared to 9% for acoustic and 11% for haptic), especially at posterior regions (50% increase in alpha at O2 and Oz for visual imagery attention versus visual external attention) and potentially gives support to the idea that alpha is involved in the gating, or suppression of sensory information which might be distracting to the imagery task at hand.
Cooper (2002) showed changes in alpha amplitude between external vs internal tasks were greater for visual attention across posterior regions of the brain.
In summary, the data seems to support the idea that increases in overall alpha activity is related to both internal and external attention in a manner that involves some inhibitory action. However there are still significant questions about the underlying alpha activity that is actually being measured in these studies. Do the differences (e.g. with studies employing the TSE technique) actually reflect differences in alpha “oscillations” or are they just reflecting a change in the overall spectral composition of the EEG waveform? Future studies which provide stronger evidence of a true oscillatory alpha components will help to clarify these uncertainties and reveal further perspective to alpha’s role in attention.