Caveat lector: This blog is where I try out new ideas. I will often be wrong, but that's the point.

Home | Personal | Entertainment | Professional | Publications | Blog

Search Archive


Voytek Neuron paper: "Dynamic Neuroplasticity after Human Prefrontal Cortex Damage"

This post is about the second of my most recent publications, "Dynamic neuroplasticity after human prefrontal cortex damage", published in the journal Neuron. This paper was—it seemed to be—the next logical extension of my recent PNAS paper (open access!), "Prefrontal cortex and basal ganglia contributions to visual working memory".

(EDIT: Here is the official press release and some media.)

While the peer-review process for the PNAS paper was genuinely a pleasant experience, this one was quite a lot more difficult. For starters, this paper was actually outright rejected from Neuron. Twice. And we appealed. Twice. No one can say I'm not adamant, at least.

So as I showed in the PNAS paper, patients with unilateral prefrontal cortex (PFC) lesions do just fine when visual stimuli enter their "good" (non-lesioned) hemisphere, but they show behavioral deficits when the stimuli enter the "bad" (lesioned) hemisphere. As I explained in my post on the PNAS paper, this research was an extension of work published in Nature Neuroscience by my advisor Robert Knight and colleague Francisco Barceló (both coauthors on this Neuron paper). So unilateral PFC lesions cause attention and memory deficits for contralesional visual stimuli.

However these deficits are unlike lesions to the motor cortex, for example. If you have a motor cortex lesion, chances are you'll be paralyzed on the opposite side of your body (this is called hemiparesis, since it's paralysis on only half the body). It's not like the people with unilateral PFC lesions have no working memory or attention. They're just subtly worse at these things than people without lesions. Why is there such a big difference between the lesions to the frontal lobe that affect movement and lesions than affect cognitive functions?

It's all about connectivity and distributivity!

Think about it this way: if you cut your internet cord, you're not going to be able to access google.com. The reason is that you've removed your final, most important connection to the internet. But it you cut a hundred cords at Google, you'll still be able to access google.com, because google.com isn't centralized. It's distributed across many thousands of machines. But those 100 cords weren't doing nothing, so you should be able to measure how the information in the network rerouted to get around those 100 missing connections.

That's the idea behind this paper. Sure, PFC lesions cause problems, but how are the people with those lesions doing as well as they are? I mean, if you stick 100 people in an fMRI machine and have them do a visual attention or working memory task, their PFC will "light-up" in a task-dependent manner. So why can people with huge amounts of damage to that region still do the task? This kind of indicates that hey, maybe cognition isn't so nicely localizable.

So we built off previous work on language and motor recovery that showed that recovery was associated with task-specific increases in the homologous brain regions in the non-damaged hemisphere.

So I re-analyzed the data from my PNAS paper, as well as some older PFC attention data from our lab.

Bradley Voytek Neuron 2010 Dynamic neuroplasticity after human prefrontal cortex damage

Just like the figure from my PNAS paper post, the figure above shows the average of the two patient groups where the color represents the number of patients with a lesion in that exact brain area.

Bradley Voytek Neuron 2010 Dynamic neuroplasticity after human prefrontal cortex damage

As you can see in the figure above, when we increased the memory load for our subjects with PFC lesions they showed increasing activity over the undamaged PFC. The greater the memory load, the harder the intact PFC seemed to be working.

What was new in our study was that we showed that this compensation occurs for cognitive tasks, and it occurs very rapidly (within 600 milliseconds) and only as needed. That is, the intact PFC seems to be "recruited" when the task is too hard.

Bradley Voytek Neuron 2010 Dynamic neuroplasticity after human prefrontal cortex damage

We tried to show this "recruitment" using a very rudimentary connectivity analysis where we showed that early activity in the visual cortex (green circles at the back of the brain) was strongly correlated with activity over the intact PFC only when the lesioned hemisphere was challenged.

This work was financially supported by the (sadly defunct) American Psychological Association Diversity Program in Neuroscience grant 5-T32-MH18882 (to B.V.) and the National Institute of Neurological Disorders and Stroke grants NS21135, NS21135-22S1, and PO40813 (to B.V. and R.T.K.).

Voytek B, Davis M, Yago E, Barcelo F, Vogel E, and Knight RT (2010). Dynamic Neuroplasticity after Human Prefrontal Cortex Damage. Neuron.


  1. Mmm, very interesting.

    I haven't read your paper yet so forgive me if this is a silly question, but how long does it take patients to gain the ability to recruit the contralateral PFC?

    Can they do that from Day 1 after the stroke, or does it take a few weeks to "get the hang of it"?

    That would seem to be relevant to the question of whether the PFC is always able to do this, or whether it's normally lateralized and "rewires" itself after a stroke...

  2. We looked at both patients with left and right hemisphere lesions--we didn't have enough power (i.e., subjects) to look at left vs. right hemisphere PFC stroke.

    As for the timing, all of our patients were chronic, not acute, meaning they were tested at least 6 months after their strokes.