Neural Networking and Damage Recovery

Nov 03 2010 Published by under Behavioral Neuro, Neuroanatomy, Neuroscience

So when Sci got the press release for this paper, it said "PHANTOM IMAGES STORED IN FLEXIBLE NETWORK THROUGHOUT THE BRAIN". I went "wut?!" and read ahead. Those press releases, how they do lie exaggerate. But this paper is cool and well worth the blogging, so I got myself a tidy little copy and settled in to read.

When people hear about recovery from brain injury, what they usually think of is something like a concussion, or a car accident, or something else traumatic and involving your head being impacted by something of magnitude.

But what we are actually usually talking about, when we talk about recovery from brain injury, is localized sudden brain injury, often caused by stroke. Voytek et al. "Dynamic Neuroplasticity after Human Prefrontal Cortex Damage" Neuron, 2010.

We've known for a while that the brain can sometimes compensate, to some extent, for damage, by shifting nearby networks (or simply the other half of the network in the brain) to take over the function for the damaged area.

For example, think of a stroke in your sensory cortex, which leaves you with no feeling or motion in one hand. Over time, your brain may compensate somewhat for the damage, with nearby sensory areas taking over, and with the hand regaining some feeling or motion. Interestingly, this works the other way as well. A famous example (From this book, which I cannot recommend highly enough) of an amputee who was missing a hand kept having feelings IN the missing hand. It turned out the sensory neurons for that missing hand had remapped onto the patient's FACE.

But anyway, what the authors were interested in here was HOW the cortex might go about compensating after injury, and at what time points that compensation happens. So they took a bunch of patients with brain injury specifically in one side of the prefrontal cortex, due to stroke. These patients suffer attentional and memory deficits. So they wanted to test memory in these patients, while ALSO looking at their brains to see what exactly was doing the remembering.

Now, you might think that this would be an fMRI study. Not so. Though fMRI is useful in terms of getting really pretty pictures and really good anatomy, what it's very bad for is TIME. A patient has to sit still for a good bit of time in order to get baseline recordings and in response to change. So while this would be good for things like, say, studies of drug effect or pain, for the quick responses they were looking for for this memory task, they needed a technique that could keep up. So in this case they used electroencephalography (EEG), a technique which measures the electrical activity of your brain. As you might know by now, your neurons send information via electrical impulses called action potentials, and these generate faint electrical readings which the EEG can pick up. EEG can pick up very rapid changes, so was ideal for this study.

They took their stroke victims (and controls), and had them complete a memory task. There were two memory tasks, one that was a low load, and one that was high. For the low load, the patients were given a series of squares, all of one or two colors, and told to remember that while they saw a bunch of random information. Then they were tested on whether they remembered. For the high load, the patients had to remember THREE colors instead of one. For both of these experiments, they made sure that the squares were only presented in one HALF of the visual field, so they tested both the half with the lesion from the stroke, and the healthy half of the brain.

On the left you can see the stroke patients, and on the right are the controls. You can see the controls on the right respond to the memory test the same way, with the same wave pattern, whether there are 1 or 3 things to remember. But for the controls, when the lesioned area was challenged with the memory task, they got STRONGER wave patterns in the UNLESIONED half of their brains, which suggests that the unlesioned half is making up for the deficiencies of the lesioned half.

They did do more than this, but I'm just sticking with the main finding, because the rest of the stuff gets into theta waves and network strengths, and my typing fingers have not the energy to explain that right now.

But what's cool about this is not THAT it happened (because people have seen stuff like this before), but how FAST it happened. The neural network that was compensating for the damage was working just as fast as if the brain hadn't been damaged at all! The brain adapted itself following exposure to the challenge, which shows that even a damaged brain is a lot more flexible than we thought before.

So how do they think this is happening? The authors hypothesize that the information is coming in on the damaged half and being transferring via the corpus callosum to the uninjured half. While I think this idea has a lot of merit, it ALSO still required some of the injured half to be functional. Are there more severe lesions where this kind of recovery simply isn't possible?

And what does it all mean? Well, the press release talked about how this could open up new ways to retrain people after stroke...but I don't know what that means or even how it would work. They talk about targeting areas rather than "training the whole brain", but...what would that even entail? It's not like we can just have the PFC lift weights, you know. When it comes to things like visual attention and memory recall, large parts of the brain have to pull together. So I'm not sure what the press release meant there.

But what I think it can mean is that it provides us with places and TIMES to look for changes in neuronal plasticity. The brain continues to show us that it's much more adaptable than we first thought, and that recovery can be just across the callosum, at least in some cases.

Voytek B, Davis M, Yago E, Barceló F, Vogel EK, & Knight RT (2010). Dynamic neuroplasticity after human prefrontal cortex damage. Neuron, 68 (3), 401-8 PMID: 21040843

No responses yet

Leave a Reply